Heivol
← All articles

A web edition is not a strategy

Strategy · Dwayne Donaldson · September 20, 2026 · 8 min read

Illustration of two tracks leaving the same starting point, the arrival of the internet. One track adds a web edition and banner ads next to an unchanged newspaper. The other moves the classified ads into a marketplace for jobs, cars and homes.

In the late nineties every newspaper had to have a website, and nearly every newspaper got one. I was unable to notice this at the time, as I had a ton of long-running KPIs I was trying to reach, like learning to walk and ride a bike, among other things. The newspapers I find interesting are the few that asked what the Internet did to the thing they were actually selling, because the answer had very little to do with news.

The common response was the web edition. Take today's paper, slap the articles on a site, slap banner ads next to them, for revenue, and set up a new department to run it. It was adopted because it had to be. The board was asking because the competitor across town had one, and "Internet strategy" needed something underneath it on the slide. Nothing about what the company sold, or to whom, had moved. Anything seeming familiar about today? - You might already be onto what I am alluding to.

I am choosing an American source, because I found that first and it blatantly supports my biased message: In the United States classified ads brought in $19.6 billion in 2000, about 40 per cent of the industry's advertising revenue. Ads for cars, jobs, homes, you name it, were all part of that revenue. By 2012 that figure was $4.6 billion. While I am no expert, I do not believe it was about the quality of the journalism. It is just that a searchable database is a better way to find a banged up Volvo for your teenage kid, who was really hoping for a Supra. Once one existed the ads moved to it and left the newspaper behind.

The one that moved

Schibsted is an interesting case. Before diving into this, I just knew them as 'that company that bought DBA.dk'. It's a Norwegian publisher founded in 1839, owner of Aftenposten and VG and, in Sweden, Aftonbladet and Svenska Dagbladet. In 2000 it took the classifieds out of its papers' hands and put them into a separate company, Finn.no, with a group of regional newspapers as co-owners. Schibsted held 62 per cent. Finn's job was to be the place Norwegians looked for a flat or a job, and doing it well meant taking those ads away from Aftenposten, the parent's own flagship.

While you can read it in their 2001 annual report yourself, I'll sum up the cool part of it: both tech strategy responses are in it. The online newspapers posted an operating loss of NOK 171 million, up from 87 million the year before. Aftenposten's advertising revenue fell 9 per cent. And Finn.no had, in the meantime, become the largest online classifieds marketplace in Norway for property, jobs and cars.

Side-by-side comparison. Adopt: the web edition, with a new media department, the paper's articles on a website and banner ads; what it sells is unchanged. Redefine: the marketplace, with classifieds moved into their own company, searchable listings for jobs, cars and homes, competing with the parent's own print ads; print classified revenue shrinks on purpose.
The same technology, two responses. Only the right-hand one changes what the company sells.

Then it kept going. In December 2003 it bought Blocket, a Swedish marketplace for used cars and everything else, for SEK 183 million, through a company controlled by Aftonbladet. In December 2023 Schibsted announced it would sell its news media to the Tinius Trust. The final agreement in March 2024 put the price at NOK 6.3 billion, and the sale closed later that year. In May 2025 what was left took a new name. The company that listed on the Oslo exchange in 1992 as a newspaper publisher is now called 'Vend Marketplaces', and it doesn't own a newspaper. As mentioned, they own DBA, so if you've bought a bicycle from there, that you later sold because you overestimated how much you would use it, you were a customer not once, but twice.. not that I'm speaking from experience.

Circling back, the newspapers are fine, for what it's worth. Aftenposten and VG still come out, owned by the trust that was already Schibsted's largest shareholder, and I don't read the story as journalism losing. I also don't know how deliberate any of it was in 2000. From the outside, twenty-five years later, it looks like a plan. I'd guess that on the inside it was an argument the classifieds people just happened to win.

Two flavours

When a technology shifts underneath a business, the strategy that comes out tends to be one of two kinds. Most of the time it's a scramble. Something has to exist, so a team is formed and the new technology gets attached to the side of the company, where it can complement what's there or defend it. That is quick and cheap, and it leaves every assumption about how the company makes money exactly where it was, which is most of its appeal. Often it's also the right call, and I'll come back to that.

Redefinition is the rare kind. It starts by accepting that the way customers get the job done has changed, and works backwards into what the company sells and what it stops doing. That puts somebody's existing revenue at risk, usually the revenue of the most senior person in the room. But if the shift really is fundamental, and you latch it onto the business on the assumption that it will complement it, you will most likely fall behind. You end up with a well-run web edition of a paper whose income has left by another door.

Owning a piece of it

Closer to home there's a case that looks like Schibsted's on paper and ended differently. I live in Denmark, where the digital letter won so completely that the paper one was shut down. e-Boks launched in 2001, built by the IT firm KMD together with Post Danmark, which later co-owned it with the payments company that became Nets. Since 1 November 2014 everyone aged 15 and over has had to receive mail from public authorities digitally. Letter volume has fallen by more than 90 per cent since 2000, and PostNord delivered its last letter on 30 December 2025. The service had been running since 1624.

Timeline of the Danish case: 2001 e-Boks launches, built by KMD with Post Danmark. 2009 Post Danmark and PBS take over KMD's share. 2012 the Public Digital Post act passes. 1 November 2014 digital post becomes mandatory for residents aged 15 and over. 6 March 2025 PostNord announces the end of letters. 4 July 2025 e-Boks is sold to CataCap. 30 December 2025 the last letter is delivered.
Twenty-four years from the first digital mailbox to the last paper letter.

So Post Danmark was in on the replacement for its own product from the first day, which is more than most newspapers could say. The setup even resembles Finn: a separate company, co-owned, doing digitally what the parent did physically. Where they part is in where the weight went. Schibsted moved its money and eventually its name over to the marketplaces. e-Boks stayed a shareholding next to a post office that remained organised around vans and sorting centres. In 2025, the same year it gave up letters, PostNord sold its share to the private equity fund CataCap. That deal closed on 4 July.

I won't pretend the alternative was easy for a state-owned operator with a duty to reach every address in the country. But "they missed it" is the wrong reading. They were in on the new thing from 2001, and the old business was never rebuilt around it.

PostNord is still here, to be fair. It delivers parcels, and parcels are a business the Internet handed to postal services at about the same time as it took the letters away.

Most shifts are small ones

Most new technologies are complements, and treating them that way is correct. No newspaper needed to rebuild itself around desktop publishing, and no postal service around barcode scanners.

Telling which kind you're looking at while it's still early is the hard part. The question I find most useful is whether the technology changes how you deliver the thing the customer pays for, or whether they still need that thing from you at all. A barcode scanner changes delivery. A searchable database of used cars removes the advertiser's need for a newspaper.

A smaller tell, and this one is a bias of mine: If you aren't an AI or AI related company, I distrust any strategy with the technology's name in it. Internet strategy, mobile strategy, cloud strategy, AI strategy. The name gives away that the technology is being handled as a topic next to the business, with its own owner and its own budget. I doubt anyone at Finn thought of their work as an Internet strategy. It was the classifieds business, run by people who took the Internet as given.

The AI department

With language models the same two flavours are on offer again, and the first one is winning by a wide margin. I personally remember sitting in a meeting with a CTO of a smaller company saying that they 'do not know how it will provide value, but they're investing everything they can into it'.

You can generally list the scramble version from memory: an AI team, an "AI strategy" in the annual report, a chatbot on the support page and a summarise button in the product. I've shipped a chatbot myself, so I say this with some sympathy. It was worth building and it answered a lot of questions. But the product was the same after launch, which is fine if a complement is what the shift calls for.

Last thoughts

Latching on is a fair choice, and most of the time it's the right one. Most new technology really is a complement, and those web editions were worth building.

What I'd like you to take from this is a pair of questions. The first is the one the new media departments were handed: how do we get what we already sell onto the new technology. The second is the one Finn looks like an answer to: what do our customers still need from us, now that the technology exists. Same company, same year, and only the second question could have told Schibsted that the thing it sold had changed.

With the latest shift, both questions are on the table again. Which one is your AI work answering, and has anybody asked the other.

Sources

If you read either case differently, particularly if you were there, write to [email protected]. This is also the kind of question we work on in consulting engagements.