The subject of an e-mail chain that went around the INMA initiative leads this week was: “It’s official — open Web is in rapid decline. Google says so.” 

This was clearly tongue in cheek. Documents leaked from Google’s anti-trust cases that showed their argument. And we had a field day arguing all sides of it. The full filing is here for those who want to read it.  

My initial response was this: It’s exactly what we’ve been saying for a year or so — the Web is in decline. We don’t know how quickly, but this transformation is huge and it’s already started. 

One of my colleagues disagrees (and had the data to prove it), but we need to break it down into past data and what we know is coming.

Firstly, the Google argument:

They are arguing the open Web is in significant decline, mostly due to other companies, so how could they possibly be a monopoly (Hard Fork podcast argues this well). And it’s fair to say the argument made by Google is self-serving, but it doesn’t mean they are wrong.

The question is whether the debate is really about advertising (as my colleague Gabriel is likely to cover in the advertising initiative here) or search, discovery, and consumer behaviour. I will focus on the latter.

At our recent CEO Roundtable, many people told us ChatGPT had replaced search for them. Answer engines are certainly a go-to for many people. I noticed that in the Comet browser, consumers have the option to use Perplexity or Google. We see this across a number of places now. 

The fact is: There is a viable alternative and people are starting to use it.

However if we look at “time spent,” panel data suggests time outside the top walled gardens hasn’t fallen off a cliff and, in some cuts, has grown since 2017. The decline is relative power and control, not instantaneous disappearance.

For news publishers, relative decline is what changes strategy.

Why this still adds up to a “post-traffic era”:

• Answer engines & AI browsers (plus Meta/TikTok/Amazon ecosystems) are siphoning intent and habit from the open Web. Less referral, more in-stream answers. That erodes the economic engine that made “free content + advertising” work. This is going to be a tough, if not impossible, business model within a few years. 

• Attribution opacity increases. Brand recall decays when content is consumed as an answer, not a destination. This is a serious challenge for news organisations and one we need to spend time thinking about. 

• Adtech concentration means the incremental efficiency gains accrue to platforms faster than to independent sites. This will be the same with commerce.

So, no the open Web isn’t dead (yet). But the business model is starting to look that way. And no business viability makes it fragile in the long term.

What publishers should do next (and what we’ve been advising at INMA):

• Shift from traffic to customers. Re-anchor KPIs from UPVs/time-on-site to propensity, habit, LTV, and segment health. Kill vanity metrics.
• Strengthen direct relationships. Logged-in users, newsletters, apps, memberships, communities — anything that builds repeatable touchpoints that you own. See takeaways from our Building Direct Audience master class on moving from traffic to direct audience, habit and  community.

• Diversify revenue. Subscriptions, B2B/licensing, events, education, premium utilities (crosswords, games), commerce with clear editorial guardrails.

• Invest in first-party data and consent UX. Treat identity and preference data like crown jewels. Make value exchange obvious and fair. 

• Consider your move from SEO to “AEO” (Answer Engine Optimisation): Understand whether you want to appear in LLMs and, if so, how you package all or some of your content (structured summaries, provenance, freshness, entity clarity) while defending brand credit.

• Negotiate from strength. Explore licensing and APIs with AI platforms — but set red lines: attribution, logs, usage bounds, and a real economic model. Use scraper measurement and blockers; more on that from INMA here.

• Reduce programmatic dependence. Tilt toward direct, programmatic-guaranteed, and high-attention formats that prove outcomes for advertisers.

• Ship product improvements continuously. Frictionless onboarding, simplified offers, value-forward paywalls, and habit loops (alerts, streaks, utility).

Nuances to remember:

• Usage beyond walled gardens can be stable while economic leverage migrates to aggregators and AI interfaces.

• The pace is uncertain, but the direction isn’t. The cliff isn’t tomorrow, but it is real. AI browsers are only going to speed this up, and now Google is in that race, too. Strategy should reflect compounding effects already visible in referral, search behaviours, and ad dispersion.

• Context matters. Some premium brands will hold pricing power, the long tail may well feel the squeeze first.

The open Web hasn’t flatlined. But as a stand-alone engine for journalism that is often funded by programmatic, it’s almost certainly past its peak. We’re in the post-traffic era, which means focus is needed on owned relationships, durable value, and product excellence. The companies that act now will control their destiny when the answer — not the clicks — become the default.

If you’d like to subscribe to my bi-weekly newsletter, INMA members can do so here.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .