Future PLC's 67% Profit Drop Puts a Public Number on the AI Overviews Tax
Future PLC reported a 67% drop in pre-tax profit for the six months to March 31, 2026, with website sessions down 15% and Google Search and Discover traffic both off roughly 20% year over year. The £4 billion market capitalization the group held in December 2022 has shrunk to about £280 million. Around 60% of group revenue still depends on Google as a traffic source.
A public-company earnings call is doing what survey decks couldn't
For two years the AI Overviews story has been carried by survey data, third-party traffic studies, and one-off publisher complaints. The numbers wandered between 10% and 60% depending on whose methodology you trusted. Future's H1 2026 results are the first time a publicly listed publisher network with 200-plus brands has filed an audited number on what the search shift is actually costing.
Revenue fell to £349.1m from £378.4m. Adjusted EBITDA dropped 24% to £83.3m. The operating margin halved, from 18% to 9%. PPC Land's breakdown of the segment data shows the damage is concentrated in the lines that depend most on session volume: eCommerce affiliates and UK and US programmatic. Shares fell as much as 28% intraday before recovering to close up 11% on the buyback announcement, which tells you most of the market had already priced in the bad news, just not this much of it.
The line that should be circulated in every CMO deck this quarter came from CFO Sharjeel Suleman: "We previously assumed audiences would stabilize via Discover or Search. Instead of stabilization, we're assuming continued decline." Internal forecasting at Future has officially moved off the "wait it out" framing. That shift, more than any single revenue line, is what the equity market reacted to.
The four-bucket segmentation is the audit framework everyone can borrow
CEO Kevin Li Ying split the portfolio into four buckets that are worth stealing as a portfolio audit for any marketing team that owns more than one content property or affiliate revenue line.
Destination brands sit at 9% of revenue and grew 5% in the half. These are properties with direct traffic, app installs, and audience habits independent of Google. In Future's portfolio that's titles like The Week and a few of the gaming and tech brands with cult readership.
Brands in transition make up 45% of revenue, down 5%. These are the ones being actively rebuilt for a post-search world but haven't crossed the threshold yet. Marie Claire and TechRadar sit in this bucket, according to Press Gazette's coverage of the call.
Non-diversified brands are the painful 15% of revenue that declined 18% in the half. These are still acting like 2019. They will get sold, sunset, or restructured. From what I've seen, an 18% decline rate seems to be the line where a public company runs out of patience.
Portfolio brands round out at 31% of revenue, down 7%. Lower priority for diversification investment.
The reason this matters for anyone not running a publisher network: this is the cleanest public taxonomy of what Google dependency actually looks like on a balance sheet. If you mapped your own content properties or your client's brands onto these four buckets, what percentage would land in "non-diversified, declining 18%"? That's the number worth knowing before the next planning cycle.
What 60% Google dependency does to a P&L
The eCommerce affiliate line dropped 24% organically to £32.3m. UK programmatic advertising fell 19%. US programmatic fell 16%. The pattern is consistent across every revenue line that depends on a visitor reading a product comparison or buying guide page, and the hits compound because programmatic CPMs drop with session volume.
This is the math worth internalizing. A 15% session decline doesn't produce a 15% revenue decline. It produces a 24% revenue decline in eCommerce affiliates because the page views that survive are skewed lower-intent (the buying-guide visitors are the ones Google now answers in the AI Overview). And it produces a 19% programmatic decline because the sessions that remain monetize at lower CPMs without the comparison-content premium.
The 20% audience decline in H1 accelerated from a 10% decline disclosed in September 2025. The trajectory is steepening, not stabilizing. That's the part the equity market actually punished. The earnings miss was a survivable number on its own. The acceleration was not.
For comparison, Condé Nast restructured earlier this year to operate at 25% Google dependency rather than waiting for search referrals to return. Future is still at 60%. That gap is roughly the cost of being publicly traded and structurally optimized for the old model. Private companies can move; quarterly-reporting ones get punished while they try.
Future Optic and Helix: the bet most publishers will copy
Future booked £2m of revenue from Future Optic in H1, with £10m expected for the full year. The product helps brands track and influence how AI engines surface them. Helix, a first-party data engine that launched in March 2026, delivered a 21% CTR improvement on Future-served impressions versus non-Helix impressions.
These are the right bets. They are also the same bets Microsoft just made free in Clarity, which is going to compress the margin on Future Optic faster than the £10m booking suggests. The first-party data engine play is the same one every publisher with a logged-in audience is now scrambling on. The differentiation window seems narrow.
Future+ membership crossed 200,000 paid subscribers across 7 brands. That's a real number, but at a back-of-envelope £30-£60 per year, it's somewhere between £6m and £12m of recurring revenue. Useful, not transformative. The destination-brand category is being built one membership at a time and it is slow.
The audit to run before the next planning cycle
Pick a 90-minute window this week and do the four-bucket sort on your own content and affiliate properties. For each property or revenue line, score three things:
- What share of sessions come from Google Search and Discover? Pull the last 90 days from GA4 or Search Console.
- What share of revenue maps to those sessions? This is the multiplier. Affiliate and programmatic land at a higher multiplier than newsletter-driven or direct revenue.
- How fast is the Google-dependent share declining quarter over quarter? Trailing three quarters is enough to see the slope.
Anything in the "Google share above 60%, declining quarter over quarter, no direct audience play in development" bucket is your non-diversified-brands equivalent. Future's market is pricing those at a 67% profit-drop premium. Yours probably will too, just on a slower clock.
The brutal read on Future is that the company saw the AI Overviews shift coming, raised the alarm publicly, started building Future Optic and Helix on the right thesis, and still got run over because the underlying decline is moving faster than the diversification math. Most marketing orgs aren't even pretending to do the math yet. That gap is the actual story.
The piece most teams will get wrong is assuming the bottom is in. Suleman's "continued decline" framing is the part to take at face value. Plan revenue assumptions around it, not around a recovery that hasn't been signaled by the company being run over by the same shift.
Notice Me Senpai Editorial
