“Large multinationals arrive in the country, contribute nothing in the way of paying local taxes, and exfiltrate value and data (“the new oil” as it was unironically christened by The Economist). It is essentially digital colonialism.”
The ugly face of what Christie calls digital colonialism was on show at a recent industry event. A handful of companies had speaking slots.
Long-term focus
Local firms spoke about serving small business, building skills and capability. Their focus was longer-term.
Meanwhile two of the multinationals that got to speak made short term sales pitches. One even used the occasion to push its latest promotion.
“…there are other approaches. Ones that involve paying taxes that provide for schools and hospitals, keeping data onshore and respecting te ao Māori, acknowledging the value of New Zealanders’ privacy, and building a resilient digital sector that will provide fulfilling, high-value jobs for Kiwis for decades to come.”
Taxes
Paying local taxes for digital products is a sore point. Yet it is not unusual for countries to tax foreign resources firms like miners and oil explorers.
On that basis, it makes sense to treat the ‘new oil’ the same way.
Tax on digital profits is being addressed at the international level. The process will be slow and could be unsatisfactory. Yet a small country like New Zealand would do better to fall into line with other like-minded nations and not go it alone.
Jobs
Jobs are critical. We have low unemployment today. Indeed, a halt to immigration means we are desperately short of skilled workers.
Yet we may be a lockdown away from widespread company failure and layoffs.
While multinationals use locals, and in cases pay well, much of the work is in sales or administration. The high value-add work tends to take place close to corporate headquarters.
More high value jobs means building more capability. It would give young New Zealanders better career paths. And that would seed interest in tech related subjects in schools and tertiary institutions.
If we get this right, there will be more corporate headquarters in Auckland, Wellington and Christchurch. This would be better for the wider economy.
“…Rebuilding New Zealand’s economy in the aftermath of the Covid-19 pandemic, and under the shadow of climate change, is a challenge that we have not seen since the end of World War II. The decisions we collectively make now have the potential to impact, positively or negatively, generations of Kiwis to come.”
Priority
There are ministers and opposition politicians who get this. Building digital capability is low down the priority list at the moment. If more prominent industry personalities speak out, we can push it higher up the agenda.
“We should be planning for our own data management, cyber security and artificial intelligence applications, and how these can be implemented across all of our sectors: agriculture, education, finance and others.
“Building and delivering value for the current and future generations, now that technology is interwoven into every aspect of our communities and our economy.”
It’s hard to disagree with any of this. A good place to start would be with government. Even now, government buyers appear to have a built-in reluctance to choose local technology. Fixing that would be the best place to start.
Unlike other search engines, Duck Duck Go doesn’t track your searches. You’ll see advertising based on your search terms, but they don’t relate back to earlier searches.
Nor are they based on your recent web activity elsewhere.
This is a different business model to Google which attempts to build profiles based on your activity. Google doesn’t just track your searches; its tentacles are everywhere. By some estimates three-quarters of all websites report your habits back to Google.
Stalker
This explains why some advertisements stalk you as you navigate the web. It can be surreal.
While a lot of people don’t care about privacy in this way, others are concerned.
The vast amounts of data Google collects are enough to identify an individual. Thanks to the ability to read most emails, Google knows where you live, what you do and can make assumptions about how much money you earn, what you spend and who you vote for.
Google reckons
Away from privacy, this approach has another advantage. Because Google thinks it knows about you and what you want, it uses your profile to send customised search results your way.
T
his can be useful. It can also be a problem. It means Google searches are not neutral. If two people search for a certain term, the may not both get the same answers.
This isn’t always helpful. You might want the best quality information, not what Google think’s you’d like to see. There’s no way of knowing that Google’s filters give you the best. With Duck Duck Go everyone would see the same result.
Duck Duck Go tricks
The search engine has a couple of help tricks up its sleeve. Let’s say you want to know more about someone you meet on Twitter. Type their address into the search bar and you get their profile.
If there is a weakness, sometimes there is not enough depth of coverage. In particular, it doesn’t do a great job of finding New Zealand-specific material.
This hasn’t changed, or if it has changed, it hasn’t changed enough. It can still be frustrating to use at times. You may need to switch back to Google to handle a specific search.
Away from New Zealand searches, Duck Duck Go does well enough. It is better than before.
Google often seems to be more interested in delivering users to sales outlets than information. Duck Duck Go doesn’t have a news filter, so a search can mean wading through lots of sales sites to find more independent information. It would be great if a news search was an option.
Bang Bing
What the search engine does have is something called bangs. This is a shorthand way of restricting a search to a single site or organisation. So, to look on Bloomberg for information about SDNs, type:
!blmb software defined networks
This doesn’t always work. The search above drew a blank. Trying the same search using The Economist bang, the browser couldn’t open anything, not even a 404 page.
Duck Duck Go still isn’t the best choice for most searches, but it is a more private choice.
Traditional laptops, that generally means the low cost models sold to families with school age children, look and feel dated next to modern MacBooks and Surfaces.
This observation hints at something deeper going on behind the scenes.
Premium mobile computers typically include technology that was originally designed to be used in mobile phones. The M1 processor used in today’s MacBooks derives from an ARM chip Apple developed for the iPhone.
Microsoft uses another type of phone derived ARM processor in its latest Surface Pro models.
Power-efficient ARM processors
Compared with the Intel processors used in more traditional laptops, ARM sips power. Computers made with ARM can go the best part of a day between charges. The M1 MacBook Air battery gets close to 24 hours.
Huawei’s MateBook also incorporates phone-derived tech in a sleek laptop form. It’s no accident that Huawei is a phone maker bringing its expertise to the laptop market.
There are a 2-in-1 and similar devices from HP and Lenovo. While they might not derive directly from phones and may include Intel processors, they mange to have many phone-like characteristics.
Legacy laptop design
In contrast, Dynabook and the other more traditional computer designs trace their ancestry direct from flip-lid laptops. It’s a format that has been around since the mid-1980s.
Yes, the Dynabook is slimmer than those models. It is way more powerful and its batteries last longer. It is better. But its pedigree comes from the old breed. Not from the new phone lineage.
Where phones become PCs
Phones and PCs have been converging for more than two decades—especially as PC sales waned and smartphones soared. Lockdown-driven work-from-home trends further blurred the lines.
Today, there are far more phones in use than PCs, and for many—even those who own both—the phone has become the primary computer.
Creative tasks still favour PCs
For creative work, like editing a movie or drafting a novel, computers still pull ahead. Sure, you could do it on a phone, but a big screen and keyboard make a world of difference.
Meanwhile, devices like tablets increasingly mimic phones—often with SIM slots—making them feel more like oversized smartphones.
While tablets are not designed for voice calls, that’s no longer a phone’s primary function.
Always-on, everywhere connectivity
In an era of ubiquitous 5G and abundant wireless bandwidth, it’s hard to remember life without constant internet access.
Apple blurs device boundaries by using ARM across iPhones, iPads and MacBooks—making their tech stack remarkably uniform.
Microsoft has struggled with ARM compatibility for Windows apps, since many haven’t been rewritten to suit the architecture.
Future Windows releases may improve this, but Windows 11 already supports running Android apps (i.e. phone-made apps).
Apple’s new Macs do the same, running iPhone apps natively. The convergence is well underway.
ARM chips leap ahead
Arm processors are at least a generation ahead of anything Intel has. The traditional chip maker is in a tailspin and does not have a plausible roadmap.
At the high end, MacBooks and Surface devices dominate. At the other end, Chromebooks—essentially cloud-driven laptops—offer simplicity in a modern form.
Chromebooks may be simple, but in their own way they are every bit as modern as MacBooks and Surfaces.
The internet-dependent Chromebook
There’s not much phone hardware in a Chromebook. Yet they share one important characteristic with phones. Both sets of devices need a constant internet connection to be any use. Most Chromebooks are budget devices, yet Google’s Chromebook Pixel attempted to bring premium build quality to the category.
You could work with a laptop on an internet-free desert island. A Chromebook is pointless without a connection.
Chromebooks, MacBooks, Surfaces and modern tablets embody progress in a way that legacy Windows laptops no longer can. We’ve crossed a threshold—in a few years, the shift will be clear in hindsight.
There’s a nasty example of the algorithm bias can do from the UK.
New Zealand has an algorithm charter which could protect us from similar problems. Although that’s not certain, read on.
Thanks to the Covid-19 pandemic, schools in England and Wales were closed during this year’s exam season. The British school year ends in July and the main exams are held in June.
Students couldn’t sit exams the normal way. Instead the exam authorities set up an assessment system. Like other things these days, this meant going digital and using an algorithm.
The tyranny of a normal statistical spread
The exam regulators made a point of using a system that would give a normal statistical spread of grades. That way they could avoid grade inflation.
It’s important for another reason. In the UK there is stiff competition for the best university places. They go to the students with the best exam results. The entry conditions for certain courses can be strict and tough.
T
o get exam results, the regulators used an algorithm that combined grades given by teachers with a student’s past performance and the past performance of their school as a whole.
I
n many cases, as many as 40 percent of the total, the qualifications authorities marked students down, below the grades recommended by teachers.
Take from the poor, give to the rich
There was one huge problem with the exercise. It was skewed towards giving students from the ‘better’ schools a shift up and those from the underperforming schools a penalty.
In the UK the best schools are all in the richer areas. People pay a huge premium to buy a house in a better school zone. Which means the exam results rewarded students from better off families.
The bias was huge. The Guardian newspaper described the algorithms used as “a sociological sorting process which entrenches class divides in the state system”.
’…by building in a criterion of past school performance to this year’s A-level and GCSE results, Ofqual has tied the fortunes of individual students to pre-existing inequalities of outcome.”
Algorithm bias means talent misses out
At first, many less well-off students who expected places at Oxford or Cambridge or, say, medical school missed out.
A-levels are important in the UK, to a degree they determine the next decade of a students’ life. They are more important than New Zealand’s NCEA exams in that sense.
This week the authorities backed down and went back to grading students based on teacher assessments. Which may fix matters, but after a huge amount of stress and upset plans.
New Zealand’s algorithm charter might not stop a similar abuse here but it could help. That’s because it makes algorithm decisions and the logic behind them transparent. The problem with the UK algorithm was less a lack of transparency and more a set of assumptions that are neither fair nor just.
This year a lot of people will pay NZ$2000 or more for a phone.
Apple set the tone at the end of last year with an NZ$2100 iPhone X. Now Samsung has joined the party with an NZ$2000 Galaxy Note 9.
One way to make sense of these prices is to calculate how many days you’d need to work to afford one—this puts the cost in more tangible terms than raw dollar amounts.
You can pay less. A basic iPhone X with 64GB of storage costs NZ$1800. The more expensive model has 256GB.
Samsung has an NZ$1700 Galaxy Note 9 with 128GB of storage. The NZ$2000 model comes with 512GB.
Whether you need that much storage when cloud storage is plentiful and mobile data is cheaper is beside the point.
Inflationary
These are two examples of how New Zealand’s Consumer Price Index or CPI is the nearest thing to an official measure of inflation.
In the most recent year, it was 1.5 percent. That means consumers paid 1.5 percent more for a typical basket of goods and services in the year to June 2018 than a year earlier. This contrasts sharply with the broader telecommunications sector, where the real cost of telecoms services fell 6.3% in 2017, driven by steep drops in fibre broadband pricing.
Expensive
At NZ$1700, the Samsung Galaxy Note 9 is $100 more than last year’s Note 8. That’s 6.25 percent higher: more than four times the CPI increase.
Apple’s iPhone X doesn’t have a year earlier model to compare.
Instead, we’ll look at the iPhone 7 and iPhone 8. When it launched the iPhone 7 was NZ$1200. A year later the iPhone 8 went on sale at $1250.
That’s a four percent increase. Apple’s markup is smaller than Samsung’s, but still well ahead of the CPI.
Everyone is at it
It’s not only Samsung and Apple. The prices of Huawei phone models climbed over the years.
Even Oppo, where the phone’s low price is the most important feature, has increased prices.
If anything, Huawei and Oppo’s price increases have been steeper than Samsung and Apple’s because they come off a lower base price.
But don’t phones get better
You might argue that the newer phones are better so phone makers can expect to sell them for more money. There’s something in this, see below.
Phone prices were stable during for years while annual upgrades meant huge leaps in functionality. Today’s upgrades are incremental while prices leap.
Apple shows the way
Apple has always lead the way on phone prices. It’s no accident it is the world’s biggest company and enjoys large profit margins. That trillion dollar valuation didn’t come by chance.
When it launched the iPhone X last year, Apple showed it could push phone prices above the NZ$2000 mark without denting sales. That opened the door for its rivals to charge more. They won’t admit it in public, but the iPhone acts as their benchmark.
Apple sells fewer phones than Samsung or Huawei.
The iPhone makes up around 20 percent of the handset market worldwide. It accounts for around 80 percent of profits from phone sales. Almost all the remaining profit from phone sales goes to Samsung.
Profits
It’s not clear how profitable the other main phone brands are. It’s not even clear if they are profitable. The companies don’t break out figures in the way that Apple and Samsung do. Yet it’s clear they are not making big margins.
Until a couple of years ago the Android phone market taken as a whole ran at a loss.
Things have changed. In part that’s because phone makers have pushed up handset prices ahead of inflation. It helps that some of the big names have either gone to the wall or wound down their operations.
Price rises have two sides
Inside the phone business, people talk about the average selling price or ASP.
According to IDC’s Worldwide Quarterly Mobile Phone Tracker:
…”climbing ASPs continue to dampen the growth of the overall market”
…”Consumers remain willing to pay more for premium offerings in numerous markets and they now expect their device to outlast and outperform previous generations of that device which cost considerably less a few years ago.
IDC says worldwide phone ASPs are up 10 percent in the last year.
Sharper prices lower down the market
Phone makers love to tell investors they have managed to increase the average selling price of their phones.
In some cases, they have done this by bumping up prices on their flagship models while fighting tooth and nail further down the market.
You can still get bargains. Spend NZ$500 to NZ$600 and you can end up with something great. It won’t have the latest camera or tonnes of storage, but not everyone needs those features.
High prices could be here to stay
New flagship phones are expensive to make, but the cost of building a phone is a fraction of the selling price.
Putting more lenses and more camera sensors may cost a phone maker a dozen or so dollars. OLED displays, curved glass add to costs. Perhaps the biggest extra cost is the memory chips needed to boost a phone’s storage, there is a trend towards higher storage in phones.
Higher phone prices are unlikely to go away soon. The glory days of fast rising phone sales are over.
People are now holding on to phones for longer, squeezing more value from the money they have already spent. So it becomes important for each sold phone to contribute a little more profit.
This post was originally written in November 2024.
WordPress was great until about a decade ago.
It was open source, which meant it could be free. It was simple and straightforward to use. It was customisable to the point where even non-technical users could build websites that didn’t look like WordPress sites.
WordPress’s stated philosophy was about democratising online publishing. Anyone could do it. You didn’t need skills or deep pockets. You could fire it up, get online and blog.
Unwelcome change
Then around a decade ago, the team behind WordPress began tinkering with it. It became bloated and complicated. It did what software companies often do: It added features that many users neither asked for or wanted.
This meant WordPress sites became cumbersome. There were ways of taming this process, but only up to a point.
WordPress site owners found they were being penalised by Google in search listings as their sites became slower. This was a warning of what was to come.
Gutenberg
Gutenberg was the first clear sign that WordPress was heading in the wrong direction.
For years WordPress was the best way to build a blog or a basic editorial web site.
It was straightforward. Although it needed some technical know-how, early WordPress was accessible. There was a great community of users willing to share information and help newcomers.
Online tutorials, user meetups and WordCamps filled in the gaps.
For a while it felt as if Wordpress really was democratising publishing.
The beauty of simplicity
You didn’t need programming skills to dig into early WordPress. But if you wanted to do more than the basics, there were easy-to-install plug-ins to add functionality.
Likewise there were many themes that let you change the look and feel of your site. Even non-developers could take things further by digging in to the CSS or tweaking lines of PHP code.
This simplicity meant it was easy to get information in and out of WordPress. In particular, I could take a blog post, find the underlying HTML code, then cut and paste it into another CMS for instant syndication.
Before and after Gutenberg
Gutenberg changed everything. Before Gutenberg you would write posts in an editor that resembled a text editor like IA Writer. Gutenberg replaced this with an editor that treats every element, text, images and so on, as a block.
Some people, maybe many people, find this approach more useful. It certainly works well for building more sophisticated sites, but this comes at the cost of making WordPress more complex.
It is harder to learn Gutenberg than the TinyMCE editor it replaces.
Page design
If pre-Gutenberg WordPress was like using a text editor, the current edition is more like Adobe’s InDesign. You need to take a training course to learn InDesign; mastery takes years.
Before Gutenberg, WordPress focused on creating great blog posts. Writers could concentrate on words and finding pictures to tell better stories.
Gutenberg is more concerned with layout, how things look on a web page. This can be distracting. It is an invitation to prevarication. There’s a risk it gets in the way of productivity. It certainly gets in the way of writing editorial.
Gutenberg’s gifts
Gutenberg is great for professional website builders. It gives them flexibility.
That extra layer of complexity that leaves casual WordPress users bewildered or in the cold means they can do more. It gives them ways to charge more and it creates a usability barrier which is, in itself, a commercial opportunity.
Professionals can use Gutenberg to design bigger, better sites with more features and functionality.
They can also sell custom blocks.
At the time it looked as if Wordpress was pushing Gutenberg in order to compete with companies like Wix and Squarespace which were proprietary alternatives for companies building commercially focused websites.
The newcomers were eating WordPress’s lunch in these key markets.
While the change may have made commercial sense, it sent WordPress further away from its roots as, essentially a blogging or editorial content management system.
It was no longer about publishing.
WordPress.com
About the same time WordPress.com, the commercial, hosted version of the software found more and more ways of charging users. Running a WordPress.com site went from pocket money pricing to we-could-take-the-entire-family-out-for-dinner pricing. To be fair this would depend on which options you chose to buy.
WordPress also dragged out the huge, sprawling Jetpack plug-in which was a way of getting people with open source, self-hosted Wordpress sites to pay subscriptions. A handful of important features were wrapped into Jetpack making the subscription essentially compulsory for any but the most casual, disengaged user.
This was the second sign that Wordpress was no longer the idealist democratic publishing service it once was. It was clear it was out to maximise income.
It’s hard to complain about that. We live in a commercial world. But we don’t have to like it or accept it.
I didn’t sign up for a full CMS
I didn’t want an all-singing, all-dancing content management system. For a few years I tinkered with WordPress themes and plug-ins that stripped the complexity, bloat and sluggishness from the software. Until I realised this was a losing battle.
My first website was a hand-coded, flat affair written in HTML and CSS with no database. I considered returning to this, although with 1500 posts that was a daunting prospect.
Since then I settled for Ghost and Micro.blog. The two options offer different approaches, different features. Ghost is clearly aimed at journalists, bloggers and online publishers. It is closer to the Wordpress goal of democratising publishing. Micro.blog can be used that way, but it also functions as an alternative to social media.
I didn’t see the recent WordPress controversy coming a decade ago, but by three years ago when I switched away from WordPress it was clear something like this was on the way.
As it says earlier in this post, I didn’t leave WordPress, WordPress left me.
_This post was originally published in October 2024. _
Whatever the outcome of Matt Mullenweg’s fight with WP Engine, WordPress will never be the same again.
There are two parts to WordPress. There’s the open source WordPress.org project and a commercial business called WordPress.Com which is part of Automattic.
Matt Mullenweg was the co-founder of WordPress and is the sole founder of Automattic.
Last month Mullenweg started a verbal fight with a company called WP Engine. The spat has now gone well beyond words. It has already changed the way many see WordPress and it looks set to alter the course of the web publishing software’s future.
Not necessarily for the better.
WordPress is important
WordPress is important because it is the technology behind about 40 percent of all websites on the open web. One estimate says there are a total of 65 million WordPress sites.
On the surface the fight started because Mullenweg was angry that WP Engine profits off WordPress without contributing anything.
This needs explanation.
WordPress is open source software. Open source means anyone can use the software, they can also update the code or even take the code and incorporate it into new projects. There are restrictions, but not many.
Open source and control
One point of open source software is that the initial developers can’t stop others from using it. This philosophy has deep roots—as far back as 2000, open source was seen as a way to escape vendor lock-in, particularly for countries wanting to avoid dependence on proprietary software from dominant powers.
They also can’t stop others from profiting from it, so long as they obey basic rules. There are moral obligations, but, in theory anyway, developers can’t force users to contribute money or time.
Which means demanding or expecting a contribution is just not how open source usually works. That said, there are many complications and variations on the open source theme where these things might happen.
Businesses that profit from open source project typically do contribute towards it. They might donate money or, perhaps more important, they might lend programmers or developers to work on the software.
Apparently WP Engine does not do enough of that.
Maker versus taker
This is sometimes described as a maker versus taker problem and it’s a version of something known as the Tragedy Of The Commons.
Either way, when there is an imbalance between major contributors and those who contribute minimally it harms everyone in that community. When one or more parties makes a killing off the work of others who see little reward, there are potential conflicts.
Maker-versus-taker stops some would be idealistic entrepreneurs from starting open source projects. This is sad because everyone gains from open source. It is a far better way of building applications, but it can be a better way of running things too.
Blurred lines
The difference between makers and takers is not always clearcut. But in general makers invest directly in the open source project as they grow their businesses. Meanwhile, takers are all about extracting money from the project.
Not contributing gives takers an advantage over makers.
This can discourage makers. If they see others getting rich from their work, they might take their foot off the pedal. It’s harder to be motivated when all you do is make others rich.
If takers go too far, it can give makers an incentive to flip their identity and become takers.
This isn’t quite what is happening with WordPress and WP Engine, but it sets up the background to the dispute.
Show me the money
WP Engine is owned by Silver Lake, a powerful private equity firm and it makes a lot of money. Mullenweg effectively argues it is a taker.
This is where it gets complicated. WP Engine’s biggest rival is a hosted version of WordPress called WordPress.com, which is owned by Matt Mullenweg.
In other words, Mullenweg’s fight is not just about abstract ideas of fairness, it is about market competition.
WP Engine responded to Mullenweg’s criticism by showing its contributions to the WordPress open source project. And then, for good measure, it opened legal proceedings against Automattic.
Blocked
Next Wordpress blocked WP Engine from accessing its servers. This is serious for WP Engine’s customers who need to continually update their software to keep their websites secure.
It turns out that before the fight was public, Automattic proposed a deal where WP Engine would commit to handing over 8 percent of revenue to licence the WordPress trademark. It could chose to pay money or use the money to fund its own employees working on WordPress projects.
Until now Automattic, WordPress.org and the nonprofit foundation that owns the WordPress trademark presented themselves to the world as three independent organisations. Ultimately they are all owned by Mullenweg and the distinctions between them have blurred.
His complaints about WP Engine have expanded. He accused the company of hacking Woo-Commerce, which is also owned by Automattic, as a way of collecting commissions that would have gone to Automattic. He also says WP Engine is infringing on the WordPress trademark.
Unless there is a sudden rapprochement this is going to be a long drawn out legal battle between billionaire Mullenweg and Silver Lake which has more than US$100 billion in assets.
That will be a bruising fight, but Mullenweg has more to lose than Silver Lake. Of course there is the money, but there is reputation and the moral high ground.
Who are the good guys?
For years Mullenweg has enjoyed a reputation as a leading open source entrepreneur. In the eyes of many, open source is synonymous with ‘the good guys’.
Even though he is fighting the kind of organisation most in the open source community would think of as ‘the bad guys’, Mullenweg has burnt through most of his social capital and lost the community’s kudos and respect.
No doubt the 40 percent of the web running on WordPress will carry on regardless, but it’s unlikely much good will come from this dispute.
There is a bigger concern here, the dispute has huge implications for the entire open source world. Automattic versus WP Engine could shape the future of many other projects. It would be a tragedy if all that expertise and idealism was lost as a result.
A directory of New Zealand journalists, publishers and media outlets with a Mastodon account.
This was originally posted in January 2023, so it will almost certainly be out of date, please get in touch with me if you know of any additions or changes.
Update, July 2025: The data explosion has only accelerated since this post first ran in 2007. IDC now estimates global data creation surpassed 180 zettabytes in 2025 — more than four times the 2020 total. The growth is exponential, driven by AI, surveillance, IoT and cloud services.
While cloud storage costs have fallen, so has the value of what we store. IDC estimates less than 10 percent of data is ever analysed. The rest sits in cold storage, largely forgotten.
Today, AI makes the storage problem worse. Models churn out draft documents, fake images, test runs and logs — most of which no human will ever read. It’s more data we (or our bots) save “just in case”.
An exabyte is a billion gigabytes. Which means in a year we added 800MB of data for each of the world’s 6 billion people. As much data as a 30 metre high stack of books.
It’s a lot of information.
Or maybe not. Storage experts believe that anywhere from 80 to 90 percent of stored data is anything but valuable.
Worthless storage, junk information
In 2002 I spoke to Rob Nieboer, who at the time was StorageTek’s Australian and New Zealand storage strategist. He revealed the vast bulk of data stored on company systems is worthless.
He says, “I haven’t met one person in the last three years who routinely deletes data. However, as much of 90 percent of their stored data hasn’t been accessed in months or years. According to Strategic Research, when data isn’t accessed in the 30 days after it is first stored there’s only a two percent chance it will get used later.”
At the same time companies often store many data files repeatedly in the same file system. Nieboer says it’s not unusual for a single system to hold as many as 15 separate copies of the same file.
Data storage Parkinson’s Law
According to Rosemary Stark (also interviewed in 2002 when she was Dimension Data’s national business manager for data centre solutions), storage obeys a version of Parkinson’s Law.
She said, “It’s a case of if you build it, they will come. Put together a system with 2GB of storage and pretty quickly it will fill up with data. Buy a system with 200GB of storage and that will also fill up before too long.”
Like Nieboer, Stark said there’s a huge problem with multiple copies of the same information but she estimates the volume of unused archive material to be closer to 80 percent. But she said 80 percent isn’t all junk: “It’s like the paper you keep on your desk. You don’t want it all, there may be a lot you can safely throw away but sometimes there are things you need to keep just in case you need them again later.”
Needles and haystacks
Although many companies focus on the economic cost of storing vast amounts of junk information, there’s a tendency to overlook the performance overhead imposed by unnecessary data. In simple terms, computer systems burn resources ploughing through haystacks of trash to find valuable needles of real information.
There are other inefficiencies. Stark said she has seen applications, for example databases, that use, say, 300 Terabytes of storage even though the actual data might only be 50 Terabytes. This happens when systems managers set aside capacity for anticipated needs. The situation is a little like a child’s mother buying outsize clothes on the grounds that the youngster will eventually grow into them.
Nieboer said there are inherent inefficiencies in various systems.
Mainframe disks are typically only 50 percent full. With Unix systems disks might only be 40 percent full, with Windows this falls to 30 percent.
It is a strange paradox of the modern age: we are drowning in redundant data, yet we are still struggling with the fact that vital information pages frequently disappear because we haven’t prioritised the preservation of the ‘living’ web over the storage of digital waste.
Last night I joined old friends and colleagues in raising a glass to the late Randal Jackson. It was the an appropriate send-off, something Randal would have enjoyed himself.
Over the years Randal was a rival, a colleague and a mate. Sometimes all three at once.
In the early 1990s I was working a freelance technology journalist in Wellington. There were others in town, but Randal was the most likely to turn up at the same jobs and events as me.
Often we’d be the only two journalists in the room. Depending on the time of day, we’d would repair to a bar afterwards to talk over whatever story was on offer and others besides.
It didn’t always depend on the time of day. Randal was happy to visit the bar any time.
The Randal Jackson breakfast show
At one alleged breakfast event I sat down next to Randal at 7am in a private meeting room at what was then called the Wellington ParkRoyal.
Two earnest American IT executives were there to talk about whatever overpriced product their company was trying to foist on New Zealand at the time.
About ten minutes in to the session it was clear they weren’t planning to give us breakfast. The mean swine hadn’t even organised coffee.
Randal wasn’t happy. He told them to stop. He said that in New Zealand an invitation to breakfast usually meant some kind of food and certainly meant hot coffee.
Fair enough. Apart from anything else we could smell the food and coffee in neighbouring rooms.
He turned to me, winked, then said: “I bet you didn’t have time to eat before coming in Bill?”.
It was a question. I told him he was right and that I was hungry.
Randal then said how he was also hungry, too hungry to think about difficult topics like enterprise computing on an empty stomach.
The strait-laced Americans were mortified. They looked confused and worried. Nevertheless they decided to bat on regardless.
Randal put his pen on his pocket, picked up his notebook, winked again and said: “Come on Bill let’s go and find some breakfast”.
We got up to leave.
“Now just wait”
The senior executive said something like “now just wait” then gave instructions to his junior. The younger executive left the room.
Five minutes later waiters entered with a coffee pot, a tea pot and croissants. This was more like it. The session resumed.
After another five or ten minutes a huge trolley rumbled in piled high with fresh fruit, eggs, bacon, sausages, the works. There was easily enough food for ten people.
We tucked in and listened, questioning the execs for another ten minutes before they took off for meetings. We demolished piles of food. They ate nothing. I guess they had their power breakfast before our session.
When, not long after, they stood to leave , the senior executive said if there was anything else we wanted we could just order and he would pick up the bill.
That was a bad move.
After they had gone I turned to Randal and asked: “Champagne?”. In those day fancy breakfast functions often included sparkling wine or Buck’s Fizz.
Randal said no, and ordered cognac instead. And coffee. And more of those little Danish pastries.
We didn’t get out of the ParkRoyal until lunch time, and only then because there was a horse running that afternoon and Randal needed to find a TAB.
An organisation I co-chair, NZRise, has been looking at the problem. We represent New Zealand owned digital companies who generate jobs and good incomes for tens of thousands of Kiwis. Our research shows Facebook, Google, Amazon and many other global digital companies are engaged in similar tax avoidance schemes to Apple.
Most revenues that accrue to those companies from New Zealand simply don’t get reported. They are the result of an online transaction and the money flies out of the country in the blink of an eye. No tax. No multiplier effect. No 41 per cent investment into our society.From a business owner’s perspective it also represents a huge disincentive to invest in R&D, which is already at shockingly low levels by international standards. We find ourselves at a disadvantage to our multinational competitors.Why create software and technical services in New Zealand when we will always be facing uneven tax playing field?
New Zealand has had a problem with multinational companies and transfer pricing for decades.
Yet the problem Christie writes about is on a different scale.
While the old multinational would shuffle money to minimise liabilities in New Zealand, they still paid some tax. They employed people, trained people and contributed to the economy in other ways. They funded university chairs, sports clubs and other worthy causes.
If the new breed does any of that, it’s invisible.
Little contribution
The new multinationals pay next to no New Zealand tax. They employ next to no New Zealanders. They contribute little to the economy.
Sure, you can argue that Apple products make New Zealanders more productive and that’s a positive economic contribution. The net positive economic contribution may even be greater than Apple fails to contribute in more direct ways.
That is an argument against banning or boycotting Apple products. No-one is suggesting that.
It is not an argument against taxing Apple.
After all, our roads carry Apple products to market. Our schools give people the skills people need to use Apple products. Our health system keeps Apple’s customers alive and healthy. In some cases our tax dollars buy Apple products.
Google this!
You could argue something similar about Google. Some believe Google software makes workers more productive than they would be with other software. Maybe.
Some think that Google’s activities in the advertising sector has an economic benefit. Try saying that to a New Zealand journalist or someone who works in the media.
Again, these are not arguments against taxing Google.
Google is quite happy to sell its products and services to New Zealand government departments that it doesn’t help fund.
It’s harder to argue Facebook offers any economic benefits to New Zealand. If anything it undermines productivity. It is the digital equivalent of an all-sugar diet.
Christie has a good point
There’s little chance Apple, Facebook and Google will stop selling if we force them to pull their economic weight.
Until recently the problem was limited. Most of the non-contributors were technology companies. That’s changing with services like Uber muscling in on our markets. If things continue, the giants will hollow out our economy. Let’s not allow that to happen.
It’s been said that what the companies do is legal. That’s true. It doesn’t make it right. We have the power to change that. We have left this problem in the too hard basket for too long.
While Christie’s argument is focused on tax avoidance and economic contribution, there’s a wider conversation about market dominance. We saw this in 2024 with the US Department of Justice’s major antitrust case against Apple, which specifically targeted the ‘blue bubble’ messaging lock-in as a means of unfairly suppressing phone competition.
Elon Musk paid $44 billion to take control of X-Twitter at the end of October 2022.
It’s been a ride ever since with story after story about the stupid, evil or inexplicable things that have happened. There are far too many to mention here.
Soon after the takeover there were many posts from disgruntled ex-employees and high profile internet personalities predicting imminent disaster.
Down, but far from out
Some said the site would fall over. It has done. The last year has seen several widespread outages. But there hasn’t been the persistent meltdown that was talked about.
Likewise, there were claims Twitter user numbers would plummet. As we shall see in a moment, they are down, but it’s a steady decline, not a rapid dive.
It appears that the technology and the site’s user stickiness both had more momentum than critics expected. That is about the only positive angle to this story.
Last month both Elon Musk and CEO Linda Yaccarino made public statements about Twitter usage, user numbers and everything else reaching an all-time high.
Decline and fool
A look at the figures shows something different. Twitter is declining in almost every way.
It’s relatively easy for researchers to get external data traffic data for a site like Twitter.
Highlights include an estimate that global web traffic to twitter.com was down 14 per cent year-over-year in September. It was worse for traffic to the ads.twitter.com portal for advertisers which was down 16.5 per cent.
Traffic was down 20 per cent in the US and 17.5 per cent in Australia.
Vanity press
It’s not all bad news for Musk. Traffic to his Twitter profile and posts was up 96 per cent.
Twitter is not the only social media site facing a drop off in interest. Similarweb found traffic to the top 100 social networks and online communities was down 3.7 percent. Yet TikTok was up almost 23 per cent over the same period.
Follow the money
Raw user numbers are one thing, what about the metric that matters the most: revenue?
Every month since Musk took over the year-on-year revenue has dropped. The best month was last October (2022). Twitter’s revenue was down 12 per cent that month. In December it posted a 78 per cent year on year revenue drop. The falls appear to be slowing, but the direction has not.
A conga line of startups and pre-existing alternatives hopes to profit as Twitter declines. The list includes: Bluesky, Post, Pebble, Spill, Mastodon and Threads. There are others, but these are the most direct competitors.
Bluesky mining a vein
Of these, Bluesky is the most Twitter-like. A site called twexit.nl publishes a running tally of subscriptions to Bluesky. It is the closest thing to a direct Twitter competitor.
On Monday Twexit showed Bluesky had 1,706,954 subscribers. That’s tiny compared to Musk’s claim of 500 million plus Twitter users, although few people believe his estimate.
You can see the Bluesky numbers jump every time Musk makes a move. In September the service saw a five per cent jump in sign-ups when Musk said he plans to charge all Twitter users.
It’s early days for Bluesky, potential new users need a sign-up invitation to join, although these are easy to get. The other obvious Twitter-like services, Post, Pebble and Spill have fewer members. Collectively they may add up to five million. That’s not much compared to Twitter’s numbers, even after discounting any Musk exaggeration factor.
Threads
Threads is more of a threat. It is Facebook’s Twitter alternative. Strictly speaking it is a Meta business, but the link to Facebook is strong with Threads posts now showing up on Facebook feeds.
Earlier this year Threads went from zero to more than 100 million users in less than a week. Accounts vary but most agree it does not yet have the impact you might expect from a social media site with 100 million users.
Mastodon remains interesting. It’s not a single site, there is no-one to call if you want to speak to Mastodon. It is made up of many sites or “instances”, often run by enthusiasts, bound together in a “federation”. Mastodon is an acquired taste, it is popular with certain communities and has a higher geek to non-geek ratio than other social media.
At the time of writing Mastodon’s records show it has 8.1 million users. That makes it, for now, the biggest of the smaller alternatives. Like the others, it often gets a boost in user numbers when Musk says something silly.
Here, highly exposed means between a quarter and a half of the tasks in a role can be automated using today’s AI.
Disrupted
A separate study based on LinkedIn’s Economic Graph, which identifies trends across its one billion members, found that 36 per cent of women will have their jobs disrupted by AI, compared with 26 per cent of men.
“Women tend to be over-represented in roles more susceptible to disruption by generative AI, such as medical administrative assistant and legal assistant, whereas men are over-represented in roles potentially augmented by generative AI, such as electrical and mechanical engineer”.
Recent research in Australia and reported in The Australian (not linked here because of the paywall) found:
“… 7.2 million employees – about half Australia’s total workforce – will need to re-skill and adapt to generative AI, with 3.3 million having their roles augmented by the technology while 3.9 million will face disruption. The remaining half of the workforce is expected to be largely unaffected by AI.”
You can read ‘disruption’ as meaning ‘likely to lose their jobs”.
Like the other research it found that clerical workers are in the front line, but away from rich countries that has other implications.
Historically as poorer countries develop economically, women enter the formal workforce by taking on clerical roles.
The ILO warns that one potential result of AI could be those clerical jobs that lift women out of poverty may now never emerge in lower-income countries.
Which not only makes it harder for women in those countries to enter the workforce, but it will have a huge impact on families.
It’s not entirely negative for women.
The Australian report goes on to say that more women than men have what it calls ‘soft skills’ which include communication, teamwork and adaptability.
These are skills where demand is expected to rise as more companies look to combine “people skills” with AI literacy.
Which means women are most likely to be displaced, but, on the whole those who are displaced are likely to do better than men at finding alternative jobs.
Last year the world created or replicated 64.2 zettabytes of data. The number comes from IDC, a market research firm (but the original document is no longer online).
The figure is remarkable considering three years earlier IDC was forecasting the 2020 number would be 44 zettabytes.
A zettabyte is a trillion gigabytes.
In part IDC puts the faster growth down to the Covid-19 pandemic: a “…dramatic increase in the number of people working, learning, and entertaining themselves from home.”
Ephemeral data
IDC says: “…less than 2 per cent of this new data was saved and retained into 2021 – the rest was either ephemeral (created or replicated primarily for the purpose of consumption) or temporarily cached and subsequently overwritten with newer data.”
Between now and 2025 the amount of data is set to grow at a compound annual rate of 23 percent.
The fastest growing source of data is the Internet of Things, not including surveillance video cameras. Social media is the second fastest growing source.
Growing faster than we can cope with
IDC says the amount of data generated is growing faster than our capacity to store data. The world had around 6.7 ZB of storage and that is growing at 19.2 per cent year on year.
Which means we save less and less of the generated data.
This is less of a problem than it might appear because a large fraction of data is useless. A decade ago experts found as much as 90 per cent of stored data was rubbish. It can include empty files, duplicates… or many multiple copies of identical files and temporary files that were never deleted.
Ransomware is a kind of online attack where criminals take control of data, usually company data, and demand payment to return it.
There are two main types of ransomware: crypto and locker.
The first encrypts data and files so that users can no longer read anything.
In theory you will get a key to unencrypt the files after you pay a ransom to the crooks. Locker ransomware is similar, but it typical locks down the computer so it can’t be used until the ransom is paid.
Travelex lock-down
After the Travelex attack, the company closed down the websites it operates in 30 countries. It said the move was designed to “contain the virus and protect data”.
That doesn’t quite sound right. After all, it emerged the criminals had been inside the company’s systems for the past six months. By the time of the attack there would little left to contain or protect.
The criminals say they have downloaded many gigabytes of sensitive customer data. This includes dates of birth, credit card information and (British) national insurance numbers.
News reports say the criminal gang asked Travelex to pay US$6 million at first, with the demand ratcheting up over time if it wasn’t paid quickly. It’s not clear if the company paid up.
New Zealand link
There is a New Zealand link. After the attack, the company’s branches, which include airport currency exchanges, were still providing services but were using manual processes.
Travelex is also the issuer of Air New Zealand’s OneSmart card. The card makes it easier to deal with money when overseas. It can be loaded with money in as many as eight different foreign currencies before a trip. Users can lock-in exchange rates to avoid fluctuations while they are overseas.
Air New Zealand says the card is not affected by the attack.
The company told the NZ Herald: “OneSmart does not use the Travelex foreign exchange services affected by the attack so
Onesmart cardholders are not impacted”.
Ransomware attacks declining
The Travelex attack happened at a time when ransomware incidents are falling fast. Last year the number of attacks dropped 20 per cent as online criminals turned to more lucrative alternatives.
In part the fall in ransomware attacks is because companies are doing a better job at protecting themselves.
The best approach to protection is to have data back-ups so everything ransomed can be recovered quickly. While this sounds simple, it’s something many companies struggle with and criminals know that. Among other matters companies tend to make back-ups without checking the data is recoverable.
Another problem is that a sophisticated ransomware attack can also take control of the back-ups rendering them as unusable as the main data store.
A ransomware attack amounts to a much bigger problem for the victim than the ransom demand. In many countries companies can face fines for not properly and promptly reporting an attack to the authorities.
At the same time, allowing data to be ransomed is often actionable under data protection legislation. At the least a company would need to prove it had taken due care with customer data, that’s hard to do after a ransom attack.
There’s another unpleasant twist to a ransomware attack. While the criminals often release keys after the ransom is paid, that doesn’t always happen. And in at least one reported case, the data was ransomed again by the same gang at a later date. Allowing that to happen is an open and shut case of negligence.
Responding to ransomware
If you are attacked by a ransomware gang, you may need professional help to recover data. Before you get to that stage you need to consider how to respond.
The NZ Police recommend you don’t pay the ransom. That’s understandable and makes sense if there’s a good chance of recovering the data.
Some security experts say that paying the ransom is the smartest course of action. It is often cheaper and, if you don’t have back-ups, quicker than other ways of recovering the data.
It may not work for everyone, but switching from Microsoft Word to a Markdown or text editor boosted my productivity.
Almost everything I’ve produced in the last 13 years was written first using Markdown.
If we want to be technical about it, Markdown is a simple, lightweight markup language.
At a pinch you can write Markdown using a plain text editor. It is better when you use an app. My favourite Markdown app is iA Writer.
Swiss Army knife
Microsoft Word is the writing equivalent of a Swiss Army knife. It aims to cater for every possible need.
Markdown and iA Writer are like one of those extra sharp Japanese cooking knives.
They do far less, but what they do, is done better with greater efficiency.
If you don’t know what life will bring you, the Swiss Army knife makes sense. But a chef would choose the latter to prepare a meal.
Simple, minimal, that is the whole point
The beauty of Markdown is there are a mere handful of commands to remember. There are few features.
That is a good thing. It means you can focus on writing words. Nothing else.
In this sense it is the closest thing to using a typewriter.
A few good commands
You can type out the commands for, say, bold text. That would be a couple of * symbols before and after the words you want in bold.
In a Markdown app you could also use Command-B (on a Mac) and the symbols are inserted for you. That’s the same code used in word processors like Microsoft Word.
This means there is almost nothing new to learn. You can be up and running with Markdown immediately.
Zero distraction
The advantage of this simple, minimal approach is you are not distracted by things that don’t matter.
There is no dithering over font choices or layout options.
Trust me, you can spend hours wondering if that editor waiting for your latest story prefers to get copy in Arial or Times Roman.
Faster
Simple means fast. A moment ago I fired up Microsoft Word on my state-of-the-art Apple M1 MacBook Air.
The app took three minutes to check for and download upgrades. Then it did something in the background before opening.
There are times when I have waited much longer to get started.
A Markdown editor is there immediately with a blank page ready to go.
Sure, there are times when I use Word. I have clients who expect to receive Word files or Google Docs. It can be easier to go there from the outset.
That said, converting Markdown to Word or Google Docs is no more than a mouse click away.
Comfortable
Markdown has another advantage. It is all about text.
If, like me, you can touch type, it means you can spend more time with your hands on the keyboard and less time mousing.
I find that over time Microsoft Word needs extra mouse activity – or touch screen action. That can give me overuse pains in my hands and arms. The more time you stay with the keyboard, the less discomfort.
It’s easy to miss this point, but if you find yourself cutting text from PDFs or web pages, pasting them into iA Writer is a cinch. Compare that with the fussiness that can happen when you past text into a word processor.
Markdown apps
Many of the posts on this site were written with iA Writer. A handful were written using Byword.
Byword is a Markdown Editor for Apple users. There is a Mac version and an iOS version that will also run on iPadOS.
iA Writer started life in the Apple camp. There’s a reason for that.
Markdown has a strong Apple lineage. One of the authors is John Gruber who runs Daring Fireball, a blog about Apple products and services.
Today there are Windows and Android versions of iA Writer.
iA Writer and Byline
For me two apps run on iOS, iPadOS, which for a long time was, in effect, the same as iOS, and they run on MacOS.
My first iOS version of iA Writer cost NZ$2.59 at the end of 2011.
It was, and remains, a bargain. That was the best $2.59 I ever spent on software. In 2016 iA charged a further NZ$5.99 for an upgraded app.
I’m not complaining. Even after buying MacOS apps, iA Writer works out at a fraction over one New Zealand dollar a year.
Phone, tablet, laptop, desktop
Because both apps store files in Apple’s iCloud, you can switch between Apple devices without missing a beat.
I can, and have, started writing on a phone, edited on a desktop, polished on a tablet and send from a laptop.
iA Writer and Byword are both solid apps. I recommend iA Writer over Byword because it has had more consistent attention from the developer over the years.
Although there is not a lot in it.
At the time of writing the most recent update of iA Writer was three months old. The most recent Byword was six months ago.
After a decade with iA Writer, it remains my main writing app on iPhone, computer and iPad.
There are a few minor niggles. iA Writer works best for my journalism and blog posts.
Once a story needs to go longer than a few thousand words it can be unwieldy. Last year I wrote around 4000 words for a book chapter using iA writer.
If that happens, I find it best to break the text into smaller chunks.
There is no question I’m more productive with Markdown than with any alternative. I get more done with less mental and physical strain.
Science fiction doesn’t do a great job of predicting the future. When it comes to telecommunications, it does worse.
At the time of writing, Netflix is streaming Blade Runner, a classic science fiction movie from 1982.
Blade Runner is interesting because the action is set in 2019. In other words, it is a view from almost 40 years ago of how we live today.
What did it get right and what did it get wrong? Some things are way off target. Early on, the hero, Rick Deckard, meets a policeman driving a flying car. We’re not even remotely near driving flying cars in 2019.
That’s a huge miss.
Bladerunner’s flying advertisements
Also early on, an advertisement floats overhead. Again, flying adverts are not an everyday feature of our lives. The nearest we get are banners floating behind light aircraft.
However, the advertising hoardings are giant screens. That is on the money. Large advertising screens are now a familiar sight in cities, although, thankfully, unlike in the film, they don’t project sound with their images.
Thanks to climate change, Los Angeles, the film’s setting, suffers constant rain. The writers were correct in predicting climate change, but we have heatwaves and storms, not constant downpours.
Travelling to the stars?
One flying advertisement encourages people to emigrate ‘off-world’. Travelling to the stars seems a tempting offer looking at the movie’s depiction of 2019 life.
But really? We still shoot rockets into the sky, but no-one has been back to the moon since 1972, let alone travelled the solar system or deeper into space.
And we know Blade Runner’s people travel beyond the solar system because later in the film one of the characters talks about seeing “attack ships on fire off the shoulder of Orion”.
Then we get to the main theme of the movie: bio-engineered replicants. These aren’t robots in the usual sense, but artificial humans. We are nowhere near this kind of technology in 2019.
As an aside, the film is based on a 1968 book called Do androids dream of electric sheep? Androids play a large part in life today, but they’re not human-like, they are mobile phones.
Where are the mobile phones?
This brings us to telecommunications. Where are the mobile phones that dominate 2019 life? Almost everybody in the real world has one.
It’s not as if mobile phones weren’t around in 1982. The first, albeit heavy and unwieldy models, were introduced in 1949. Motorola had practical commercial handsets in 1973.
And where is broadband or any other kind of digital service? In 1982, some homes had tele-text machines. And email started in the 1970s. I had a work email account in 1982. Sure, it was dial-up and extremely slow, but no-one in the film has anything remotely like internet access.
Blade Runner is entertaining and thought provoking, but as a foretaste of 2019 it doesn’t come close.
Technology journalists aren’t much better at prediction the future.
“Here’s the bad news: No one is coming to save you. No business is going to swoop in and provide sustainable funding for newsrooms. No new technology is going to transform the way journalism supports itself forever.
“No big, incredible deal is going to build a strong foundation for the news. There isn’t a single magic bullet that will work for everyone. Even producing groundbreaking journalism isn’t going to suddenly turn your fortunes around.”
Werdmuller says journalists need to recognise the internet is not a broadcast medium but a conversation. This is true.
Many journalists use social media as a broadcast medium. They see it as a way to draw in readers to their newspaper, radio or TV channel websites. Whether they post to X (formerly Twitter), Facebook, Linkedin or one of the newer alternatives like Threads, Bluesky and Mastodon, they often simply post a link to their stories hosted on mainstream media sites. In many cases that’s the end of their interactions.
This can be just as true in 2026 as it was five or more years ago.
Yet many New Zealand journalists have learned how to engage with readers online.
This is especially true of the journalists who have pushed out of traditional media boundaries and are no longer employed by major news brands. They may produce podcasts or YouTube video. They may publish on Substack or similar services. They may run email delivered newsletters or more traditional websites.
In the earlier 2020 story I wrote about Twitter, it could apply to any modern social media channel:
“Most use it as a broadcast medium – like a RSS feed. A number have Twitter accounts, but say little of value. Perhaps 40 percent can be said to be serious Twitter journalists.”
Change Twitter to X or any other social media channel and you’ll realise the number hasn’t changed much. Less than half of the journalists visible online use social media to its full advantage.
Social media as a conversation
What has changed is many of New Zealand’s higher profile journalists now have regular active social media conversations. Go and dig around, you’ll see many of the best-known names engaging with their audiences. It can be hard doing this among the snark and antagonism.
One innovation that I previously attempted on my site was to integrate social comments. At the time my site was hosted on WordPress. Back then I used IndieWeb tools to capture tweets or other responses to my posts on stories. I’ve did this to boost the conversational aspect of my work.
Now I have switched the site to Ghost, specially Ghost Pro because it includes the ability to create and send email newsletters. Ghost does not allow the kind of plug-ins that can be used to expand WordPress functionality.
A potential conversational superpower
Instead, it has its own superpower. Ghost publishes to something called the Fediverse.
Thanks to the Fediverse, you don’t need to be “on” a social network in order to interact with people using suitable services such as Mastodon or BlueSky.
When you publish a post on a Ghost site, it can be sent out automatically so people on Mastodon and other fediverse platforms can follow it, see it in their feeds and reply to it.
What this means from the reader’s point of view is that:
They can follow your site like they would follow a person
Your posts automatically appear in their timeline
Any replies they make to these Fediverse posts automatically come back to your site.
This means journalists don’t need to broadcast on social media. They can run their websites and newsletters. They don’t actually post to Mastodon or BlueSky. The beauty of this approach is that you are not limited to one social media service. Nor are you even limited to social media… material distributed to the Fediverse can turn up in other ways.
To date my experience using this tools has meant limited interaction. I still get more feedback via email than by the Fediverse. And I still remain active on BlueSky, Mastodon and Linkedin. There are frequent conversations, but they don’t yet all appear under the single Fediverse umbrella.
Building audience relationships
Ben Werdmuller’s story linked to at the top of this post from ends with:
“Until publishers encourage reporters and editors to engage with their audiences, they are going to miss out on the potential of Twitter.
Of course, the journalists who do this best will become media brands in their own right, which will worry the bean counters. But that’s another story…”
This continues to be true many years after those words were written even if Twitter has morphed into the dysfunctional X and other services have replace it..
Many of us who still work as journalists are now mini-brands. Publishers and editors hire journalists with a good brand. Freelancers like me get work on the back of having a brand.
This doesn’t come naturally to older journalists. We taught journalists to keep themselves out of the story. I’m still not 100 percent comfortable inserting myself into stories.
But that’s not how things work today and it definitely isn’t how blogging works.
Werdmuller has a different take on what amounts to the same idea. He writes:
“Instead of thinking in terms of having an audience, you need to think about building and serving a community. Instead of informing, you need to be listening. The opportunities to learn the nuances of your community and to serve it directly are unprecedented — but it takes work.”
It does take work. One of the skills journalists pick up is to be excellent at listening to sources. In the past we’ve not been so good at listening to our audiences. It took me a while, although judging by my earlier posts, I was onto this many years ago.
Some of the institutional knowledge that helped us understand audiences, like newspaper librarians who provided context and memory, has been lost along the way. But journalists can rebuild that connection directly.
The point here is there hasn’t been a clear dividing line between sources and audiences for many years now. Likewise, there is less of a division between journalists and audiences. We are, as Werdmuller puts it, communities. He is right when he says this takes work, but boy, it can be rewarding.
For years we wondered. What use would humanity find for cryptocurrency? Now we know. It is not necessarily a force for good.
There have been other technologies which emerged before there were practical applications.
When the first laser was built in 1960 it was impressive. Scientists thought it may one day find use in spectrometry or even nuclear fusion.
Others thought it could be used as a ‘death ray’ military weapon. it didn’t help that the Pentagon funded early research into laser applications.
In time engineers found thousands of applications. Today it powers fibre communications networks. They are used to measure distances with incredible accuracy. Application include medicine, office printers and cutting objects for manufacturers.
The killer app
When the first PC arrived, it looked like it had potential. It could do lots of things, but it did one thing very well: spreadsheets. VisiCalc, an early spreadsheet was the first computer ‘killer app’.
Likewise, the graphically gifted Macintosh computer had its power unleashed by PageMaker. It was a desktop publishing program and another killer app.
In May criminals attacked Waikato DHB demanding a ransom in return for unlocking computers.
It wasn’t the only ransomware attack that month, nor was it the biggest or most disruptive. Ireland’s health computer system was also shut down. The pipeline moving oil to the US East Coast was shut down.
All of these ransomware attacks, and most other online crimes, have a common denominator. The criminals want ransoms paid in cryptocurrency. That’s because Bitcoin and the other cryptos are harder to trace than conventional forms of money.
Ransomware and cryptocurrency
Ransomware is crypto’s killer app.
Cryptocurrency remains a shadowy world. It is not that everyone involved in cryptocurrency is a criminal. It’s more a case of every online criminal uses crypto.
For many everyday folk, their first interaction with cryptocurrency is when they need to buy it to pay a ransom.
This is not an argument to ban cryptocurrencies. Although it could be. And the stories about the vast amounts of energy needed to ‘mine’ these new currencies are also a concern.
Part of the attraction of crypto is that it remains unregulated. That has to stop. The exchanges that deal with cryptocurrency have to face the same accountabilities as other financial institutions. It has to be made harder to move unaccounted funds from crypto into traditional banks.
There is more to stifling ransomware than regulating Bitcoin and its peers. Yet the ransomware epidemic now threatens online commerce. In cases like attacks on hospitals, it is potentially a literal ‘killer app’. Regulating cryptocurrency will save lives and jobs.
Originally published May 2020. The challenge of sustaining local technology journalism in New Zealand has only intensified since then, with further media consolidation and newsroom cuts.
The story didn’t get a run in any reputable New Zealand media.
Contrast this with the extensive coverage Microsoft got the following day when it announced it was opening a New Zealand cloud region.
The Microsoft story was everywhere. It popped up at Stuff, RNZ and Reseller News among others. There were overseas runs at TechCrunch, CRN and Computerworld. The Stuff version was almost straight PR copy.
The prime minister even talked about it on TV.
Big run
The point here isn’t about New Zealand media giving the overseas company a bigger run than the local company. Although that could be a story in its own right – see Comparing the stories below.
What the contrast between two stories show is how much damage the lack of local technology news coverage does to New Zealand’s home grown technology sector.
No-one here has the resources to file a story that is, by local standards, somewhat significant.
No one is watching, does anyone care?
We no longer have a native technology press. It’s a situation which, presumably, will be worse if Stuff no longer operates as a separate entity. 2026 update: thankfully this did not happen.
The vacuum means local technology companies struggle to build awareness through traditional media channels.
Last month Bauer Media closed its New Zealand operation shutting off Peter Griffin’s excellent regular features in the Listener. The Listener has since been sold and continues publishing.
This web site is also sporadic. At the time of writing, my stories here are posted between paying journalism jobs. That means they can’t always be timely.
Filling the vacuum are many overseas sites. Whatever their merits, they are not going to zoom in on the activities of a local cloud provider.
Comparing the stories
There’s no question the arrival of a New Zealand Microsoft cloud region is the bigger news story. Microsoft is the world’s second largest cloud operator. It has many customers here and there is a pent-up demand for a world-scale cloud operator to open shop in New Zealand.
In contrast, the Catalyst story, is not much more than a feature update.
There are interesting angles to the Catalyst story. The cost of its Object Storage is on a par with costs for world scale cloud operators. It costs three cents a month to store a gigabyte.
The ‘everything stored in New Zealand’ angle is important. But it’s also an important part of Microsoft’s story. And, no doubt, Microsoft could make the same claim about only using renewable energy.
Uphill battle for local technology news
What this illustrates is a company like Catalyst struggles to be heard above the noise.
It must be galling for people at Catalyst and other New Zealand technology companies. They something innovative like introducing low cost cloud storage only to wake the following day and see a rival’s news splashed around the place.
Longer term it is a worry. Wikipedia says:
“If a tree falls in a forest and no one is around to hear it, does it make a sound?” is a philosophical thought experiment that raises questions regarding observation and perception.
Tech companies need that observation and perception. New Zealand’s tech sector no longer has either.
This isn’t just bad for technology companies—it’s bad for New Zealand. Without local technology journalism, we lose accountability, context and the ability to understand how technology shapes our economy and society. The companies doing innovative work deserve to have their stories told,and New Zealand readers deserve to hear them.
**More on journalism and media: **
This post is part of ongoing coverage about journalism business models, digital adaptation and modern reporting:
Originally written December 2010, updated September 2025, with additional context added January 2026.
In Costly Mistakes for the American Journalism Review, former newspaper reporter turned industry analyst John Morton argues that many of the steps US publishers took to counter falling advertising revenues in the run-up to the recession only made things worse. His analysis is sharp and remains a must-read for anyone following the newspaper industry.
High margins, short horizons
Morton points out that corporate publishers demanded margins of 20 percent or more from their papers. This was unsustainable. The pursuit of these returns meant cutting the very things that made newspapers valuable—experienced staff, institutional knowledge, deep local coverage.
Family-owned papers had traditionally worked on closer to 10 percent margins, still a healthy return that brought owners influence and prestige along with profit. Owning a paper was lucrative, but the greater rewards were power and status.
Corporations, however, chased higher returns. To keep margins fat, they cut investment and hollowed out editorial quality. By the time the internet arrived, publishers assumed they could leverage their media assets into digital businesses and keep the money flowing.
While huge numbers of readers visited newspaper websites, few stayed long or read deeply. Publishers struggled to understand that digital and print required different approaches, not just digital versions of print strategies.
Writing for Nieman Journalism Lab, Martin Langeveld said publishers needed to grow their online market share rather than erecting barriers. He argued that paywalls and fights with aggregators risked driving audiences away, further shrinking relevance.
Paying for loyalty
Paywalls may strengthen ties with a small core of loyal readers, but at the cost of scale. The risk is that newspapers swap mass reach for narrow influence, losing their role as central public forums.
The industry’s obsession with margins in print and its missteps online meant publishers missed opportunities to adapt. By prioritising short-term profit over long-term sustainability, newspapers undermined both their business model and their place in society.
What could have worked differently
Looking back from 2026, some alternative paths become clear:
Build reader relationships early: Publishers who invested in email newsletters, membership models and direct reader engagement in the early 2000s—before social media dominated—built sustainable audiences. Those who learned to use journalism tools to build communities fared better than those who treated websites as digital print editions.
Preserve institutional knowledge: Cutting newspaper librarians and experienced staff seemed like smart cost-cutting. It proved disastrous. New reporters couldn’t learn from veterans and investigative capacity evaporated.
Avoid the subscription trap alone:Subscription models worked for elite publications but created a second digital divide. A mix of approaches—freemium, micropayments, memberships—might have served democracy better.
None of these guaranteed survival, but they offered better odds than chasing unsustainable margins while hollowing out the product.
The loss of control wasn’t just about business model failures. As newspapers struggled, tech platforms like Google and Facebook became the primary distributors of news, treating journalism as content to drive engagement rather than as a public service.
This shift fundamentally changed who controls what news people see and how they see it. Meanwhile, local journalism withered in many markets, and ad-blocking further undermined remaining revenue streams.
More on journalism and media:This post is part of ongoing coverage about journalism business models, industry decline and missed opportunities:
Originally published July 2018. Updated January 2026 with observations on how subscription fatigue and a proliferation of paid services have reshaped the economics of paying for content.
The subscription explosion
When this post was first written in 2018, the subscription economy was already growing. By 2026, it has exploded—and the problem of limited consumer budgets has only intensified.
The average person now juggles subscriptions to streaming services (Netflix, Disney+, Apple TV+), music services (Spotify, Apple Music), cloud storage (iCloud, Google Drive, Dropbox), productivity tools (Microsoft 365, Adobe) and increasingly, news and journalism services (Substack, Patreon, individual publication subscriptions).
The core insight from 2018 remains true: people allocate a fixed budget to subscriptions, creating fierce competition among publishers, app makers and content creators.
Ben Brooks gets close to the heart of the problem with pay walls when he writes Subscription Hell. It’s hard to make money from pay walls.
The few online sites that do well from pay walls are those like New Zealand’s National Business Review or The Economist. Both serve well-heeled audiences with unique, quality content readers can’t get elsewhere.
Brooks makes two interesting points.
First, differentiation. Brooks is thinking about podcasting, but it applies to all online media. In essence he says there are thousands of undifferentiated podcasts chasing the same audience.
…but will they pay?
The implication is that no-one will pay to listen to one of the podcasts when there are dozens of free alternatives. You could say the same about most online media. This, in part, does not apply to pay wall successes like the NBR and The Economist.
Their audiences don’t have obvious alternatives.
The other point is subtle. Brooks makes the connection between people paying for apps and buying pay wall subscriptions.
On the surface these are two quite distinct markets. And yet, recently I was thinking about exactly this concept from the opposite point of view. I have a number of subscriptions to pay each month. Some are for apps or online services. Others are for, it’s a horrible word to use, but let’s go with it: content.
Pay wall, subscription software: two aspects of the same thing
When budgeting, the two are aspects of the same thing. I allow myself so many dollars a month for subscriptions. It’s a single pool of money to cover digital services like cloud storage, online music, movie downloads, pay walls and apps. What isn’t spent on apps is available for media. What isn’t spent on media can be spent on apps.
A decade ago the budget was zero. By 2018 it had grown. While it still wasn’t a huge amount of money, it was about the same as I spent on coffee. In 2026 it is well past that level. I don’t plan to let it grow higher.
I’m not alone. Two decades ago, in the mid-2000s, the budget for digital subscriptions was essentially zero for most people. By 2018 it had grown to a modest amount. By 2026, the typical household subscription spending has ballooned—one estimate suggests US households now spend over $200 monthly on subscriptions, with many unaware how much they’re actually paying.
The issue is, consciously or not, people only budget so much money for subscriptions. We have a limited pool of funds. So does everyone else. The world has a limited pool of funds for subscriptions.
On a world scale it is huge and the pool is still growing. Even so, there is not enough to go around for everyone who would like to earn money selling pay wall subscriptions or apps.
Too many sellers, too few buyers.
And there’s the problem. It’s not hopeless. Micropayment services and reader-supported platforms have evolved since 2018. Substack, Patreon, Ko-fi and Buy Me a Coffee offer ways for readers to directly support creators. I use New Zealand-based Press Patron and can recommend it.
Some publications experiment with bundled subscriptions or tiered pricing to capture different budget levels. But fundamentally, the challenge remains: too many sellers competing for limited buyer budgets. The newspapers that missed their opportunities to build early subscription habits now face even fiercer competition than they did in 2018.
Yet it’s difficult. The market for content pay walls or subscription software is not infinite.
Subscription fatigue sets in
In 2026, we’re seeing subscription fatigue. People are overwhelmed by the number of recurring charges hitting their accounts. Many have begun rotating subscriptions—subscribing for a month, binge-watching or reading, then canceling until they need it again.
This presents new challenges for publishers and creators who need predictable recurring revenue. The answer isn’t obvious, but journalists who use their core skills to build genuine communities tend to retain subscribers better than those treating it purely as a transaction.
More on journalism and media:This post is part of ongoing coverage about journalism business models, digital adaptation and the subscription economy:
Originally published March 2017. The core argument about journalism serving readers rather than industry remains as relevant today, even as the business model challenges have intensified.
At The Register Shaun Nichols writes:
“The tech press has dared to lean away from its core mission of making technology companies more profitable, says tech advocacy house ITIF.”
Long story short, it says the media moved from a positive attitude towards the industry to confrontation.
This, according to the ITIF, is because being tough on the industry makes it easier for tech media to turn a profit.
It goes on to talk about the media being ‘biased’ and distorts the public view of technology.
Yes, it’s all stuff and nonsense. There’s a lot to unpack, but here are a couple of ideas to think about.
Advertising
In the past publishers made money selling advertising to technology companies. They were a great sales conduit. It worked.
The technology industry was the tech media’s most important customer. Rivers of gold poured in.
While there are publishers who publish nice stories in return for advertising dollars, that was never a great business model. Reader are not fooled. They don’t stick around for blatant propaganda.
The advertising money didn’t buy favourable coverage, at least in the better publications. It did foster a favourable attitude towards the industry. The coverage reflected this.
The partnership also meant journalists and publishers spent time in the company of tech industry people. That too is good for creating a positive attitude.
One conclusion of the ITIF report is more advertising would repair media relations.
Readers and journalists
In the old model, advertisers paid for journalism, but journalists serve readers. Few understood this then. They still don’t.
As Nichols says, we’re not industry cheerleaders. We don’t earn cheerleader, public relations or marketing-type salaries.
Our job is to inform readers. If there is more cynicism in technology media (see the next point) then that is what readers want.
Modern reporting tools mean we know what stories rate from the minute they go online. Guess what? Readers are less likely to click on happy-slappy, isn’t everything wonderful darling stories.
In other words, journalists and publishers respond to reader demands.
Don’t shoot the messenger if they now have a darker view of the tech industry. Get your own house in order.
It’s all nonsense anyway
To argue tech media is meaner than it ways, say, thirty years ago is bonkers. The big newspapers and media sites are full of thin press release rewrites. It is common for blatant propaganda to appear as factual news. This is the opposite of what old-school journalism fundamentals taught us.
Take, for the sake of argument, Computerworld New Zealand. Thirty years ago, even a decade ago, it was breaking news stories. It was quoted in Parliament. Today, it runs nothing that didn’t start life in a public relations office.
That’s not to say all the tech media is soft. It isn’t. But the ratio of soft stories to more hard hitting news is off the scale. You have to wonder if the ITIF is paying attention.
More on journalism and media:This post is part of ongoing coverage about journalism business models, digital adaptation and modern reporting:
Originally published September 2008. At the time online media was expected to replace newspapers but not necessarily do the full job. Updated January 2026 after eighteen years when the story proved correct: online left a gaping hole.
The prediction: Online would leave a gap
In 2008, an article in the Australian newspaper: The winter of journalism’s content argued that online publishing, which was widely expected to supplant newspapers and magazines, would only go so far in replacing them and leave a gaping hole.
This worried me then. It should worry us more now.
When advertisers abandoned print media
The 2008 argument was straightforward: Advertisers were abandoning print media for online, attracted by cost-effectiveness and perceived targetability.
Yet those online advertisers preferred placing messages next to “niche interest stories”—car ads next to driving features, travel ads next to vacation content—not next to investigations of government corruption or corporate malfeasance.
Even if publishers could fund hard news, advertisers wouldn’t want it. The perverse incentive was clear: publish less accountability journalism, more marketable fluff.
However, traditionally it was those difficult, hard news stories sold printed newspapers and dragged in readers in the first place. The hard news delivered readers to the publication where they could consume the advertising.
Eighteen years later: The gap is real
By 2026, the forecast proved accurate—though not uniformly. The landscape fragmented:
**Where investigative journalism survived:
**
Elite national outlets: The New York Times and Guardian continue to invest in investigations, funded by digital subscriptions. They hire investigative teams and break major stories.
In January 2026, this list originally included the Washington Post, but that’s now questionable following recent cuts and that move illustrates the continuing vulnerability of traditional hard journalism.
Nonprofit models:ProPublica, founded in 2007, proved investigations could be foundation-funded. Local nonprofit newsrooms emerged in dozens of US cities.
We don’t see anything comparable in New Zealand where the pool of available money is not big enough to sustain even a nonprofit publication.
Specialist outlets: Sites focusing on specific niches (healthcare, environment, national security) found niche audiences willing to pay.
Where it died:
Local newspapers: The papers that held city councils, school boards and local businesses accountable have largely disappeared. Local journalism collapsed, and with it, local accountability.
Regional coverage: Mid-sized papers that once investigated identifiable regions in countries like New Zealand or the UH, US state governments, large corporations and regional issues cut investigative teams first when ad revenue collapsed.
**Specialist niches: **Coverage of courts, local government, education and civic institutions evaporated in many markets.
The gap wasn’t filled—it was papered over with press releases, wire service copy and user-generated content.
The economics that did the damage
The 2008 prediction about advertising economics proved devastatingly accurate:
Print advertising collapsed: Between 2008 and 2026, newspaper print advertising revenue fell by approximately 90 percent. Digital advertising revenue grew, but never came close to replacing what was lost.
**Online advertising favours fluff: **The algorithm-driven ad market rewards engagement and scale, not importance. Publishers learned that listicles, celebrity news and viral content generated far more ad revenue per dollar invested than months-long investigations.
The “adjacent ad” problem persisted: Advertisers still don’t want their brands next to stories about corruption, crime or corporate wrongdoing. This “brand safety” concern systematically draws money away from accountability journalism.
Investigative teams spending months on complex stories
Court reporters attending every trial
Government reporters tracking legislation - there is still some coverage, but reduced and shallower than in the past
Local government watchdogs at every city council meeting
The volume of online content exploded. The volume of accountability journalism contracted.
Alternative models that emerged
The gap wasn’t filled, but some models showed promise:
**1. Nonprofit newsrooms:
**
In the US, ProPublica, The Texas Tribune, Voice of San Diego and dozens of others proved foundation funding could sustain investigations. But this model:
Relies on philanthropic priorities (not public priorities)
Doesn’t scale to every community. There is no equivalent in New Zealand
Creates dependence on wealthy funders
2. Membership models:
Sites like The Guardian’s voluntary contributions and De Correspondent’s member-funded journalism showed readers would support quality work. But subscription fatigue limited how many outlets could pursue this. The Guardian’s needy begging is so tiresome it turns readers off what could be a useful site.
3. Hybrid models:
Public radio expanded into digital, combining listener support, foundation grants and some advertising. Both the UK’s BBC and New Zealand’s RNZ run credible online news operations. In the US, universities launched investigative centres. Some success, but not comprehensive and nothing of the sort in New Zealand.
4. Individual journalist brands:
Substack, Ghost and similar online services, let individual reporters build subscriber bases. Independent journalists broke stories—but without institutional support for legal, research and editing. In New Zealand Bernard Hickey maintains a lively news focused site with a model that sees his most important stories made available to non subscribers.
But despite all these efforts, none replaced the comprehensive accountability coverage newspapers once provided.
The democratic deficit
Here’s what society lost:
Local corruption or incompetence goes uncovered: Without reporters at city council meetings, local officials face less scrutiny. Small-scale corruption and poor governance that affects citizens' daily lives—zoning decisions, contract awards, police conduct—happens in darkness.
Corporate power unchecked: Complex investigations of corporate behaviour—wage theft, environmental violations, financial fraud—require resources few outlets can deploy. Companies know this and act accordingly. Also there is an asymmetry when it comes to access to the law, news organisations can rarely afford to defend litigation even when they are clearly in the right.
Government opacity increases: Without specialist reporters who know the territory, government press releases become “news.” Official narratives face less challenge. One phrase that comes up whenever officials are questioned on such statements is “just use the press release”.
Civic knowledge declines: Citizens can’t effectively participate in democracy if they don’t know what’s happening in their communities. The information divide became a democratic participation divide.
As foreseen in 2008: “this vicious economic cycle is nothing compared to what can happen in a society that no longer has a practical mechanism for scrutinising governments and out-of-control corporations.”
By 2026, many communities have no such mechanism.
Could it have been different?
Looking back, newspapers missed opportunities. Some alternative paths:
Earlier investment in reader relationships: Publishers who built direct subscriber relationships in the early 2000s—before social media dominated—preserved more of their audience and funding for accountability work.
Collaborative models: Had newspapers pooled resources for investigations rather than competing, they might have maintained more capacity. Some collaborations have emerged, but too late.
Public funding: Other democracies subsidise journalism more directly. The US and New Zealand largely didn’t. Whether that would have worked remains debatable, but private markets clearly failed to fund adequate accountability journalism.
Different priorities: Publishers chasing unsustainable profit margins cut investigations first. Had they prioritised accountability over quarterly returns, things could have been different.
None of these guaranteed success. But what actually happened—allowing market forces alone to determine what journalism survives—left democracy worse off.
The 2026 reality
The 2008 blog post was on the money: online media didn’t fill the gap newspapers left.
We have more content than ever. We have less accountability journalism than we need.
We have viral videos and hot takes. We have fewer reporters at government meetings.
We have elite national outlets continuing to do good work. We have local information deserts.
The gaping hole remains. It’s affecting how democracy functions. Eighteen years proved the concern valid. The question now is whether we’ll do anything about it.
There’s not much scope for optimism.
**More on journalism and media:
**
This post is part of ongoing coverage about journalism business models, democratic accountability and the information gap:
Originally published November 2019, when subscription paywalls were becoming widespread. Updated January 2026 after the predicted divide became reality.
Seven years later: The divide deepened
The 2019 warning proved accurate—and then some. By 2026, the subscription divide has become entrenched:
The information haves: Subscribe to multiple news sources (NYT, local papers, specialist publications), access quality analysis, fact-checked reporting and investigative journalism. They understand complex issues through expert coverage.
The information have-nots: Rely on free sources—social media, ad-supported clickbait and increasingly, misinformation and propaganda. Quality journalism sits behind paywalls they can’t afford, especially when subscription fatigue means even middle-class readers pick only one or two news subscriptions.
The result: Knowledge gaps correlate with wealth. Those who can afford comprehensive news coverage make better-informed decisions about health, finance, politics and daily life. Those who can’t are more vulnerable to manipulation.
Google and Facebook control almost all the world’s online advertising revenue. To get around this, news organisations and other online media use paywalls and subscriptions.
It makes perfect sense when there’s precious little advertising revenue to pay wages and other bills. Producing media costs money. Publishers learned that calling them subscriptions rather than paywalls helped, but the fundamental problem remained.
He writes: “The digital divide is about to get worse with the rise of subscription-based news media because of the failure of advertising to provide revenues for a sustainable business model.”
It’s another reason to not like Facebook. Another reason to fear Google.
Newspapers are not the only examples. Subscriptions, not advertising, pays for Video and sports streaming services. Pay-per-view is not new, but there is now more of it.
Here in New Zealand, the National Business Review hides all stories behind a paywall. The NZHerald keeps its best stories for paying subscribers. They are not alone.
A second digital divide
As an upshot, low income people who manage to jump the first digital divide and get online, come up against a second divide. Subscription costs often shut them out from the best online content.
By 2026, the economics became stark: A comprehensive news diet requires multiple subscriptions.
Local news ($10-15 a month), national coverage ($15-20 a month), specialist reporting ($10-20 a month), international news ($15-25 a month). That’s $50-80 monthly just for news—competing with Netflix, Spotify and essential software subscriptions.
Free media has stepped in to fill the gap left by newspapers. Some free sites are good. RNZ runs an excellent free news site.
Some free media is darker. People with a hidden agenda and money to spend can publish plausible looking news. Although plausibility isn’t essential here. Manipulators have free run to bombard readers with lies and misleading information.
Propaganda
Look up an international story on Google News. You’ll find links to certain sites that are openly or not so openly propaganda sites. There are Russian and Chinese examples. In some cases intelligence agencies pay the bills.
Other free news services might push extremist ideologies or misinformation. Lies are common. By 2026, AI-generated content made this worse—plausible-looking “news” sites pumping out convincing misinformation at scale, all free to access.
People who buy subscriptions end up better informed. They can make better choices. They may even live better, healthier, even happier lives than the poor souls on the wrong side of the second digital divide.
Meanwhile, local journalism collapsed in many areas. Communities without affordable local news sources became information deserts, filled by partisan blogs, social media rumours and national outlets that couldn’t cover local issues adequately.
Democracy requires informed citizens
This creates a genuine crisis for democratic societies. Informed citizenship requires access to quality information. When that access depends on wealth, democracy suffers.
Some publishers recognised this. The Guardian built a voluntary contribution model—free to read, supported by those who can pay. Some local papers offered low-income subscription rates. Major investigations often appeared outside paywalls as a public service.
But these were exceptions. Most publishers, struggling to survive, had no choice but to maximize subscription revenue. The business model challenge that drove publications toward subscriptions created unintended social consequences.
**The alternatives weren’t better: **Ad-supported models had failed, surveillance capitalism exploited readers, and direct government funding of journalism raised its own concerns. There were no easy answers.
**More on journalism and media:
**
This post is part of ongoing coverage about journalism business models, access to information and the subscription economy: