The Clicks Brief

Alphabet's second quarter was strong and the commentary around it was confident. Sundar Pichai told investors that AI Mode has passed a billion monthly users and is driving an incremental increase in search queries overall. Philipp Schindler, who runs the ads business, said Google is encouraged with monetisation performance on queries that show an AI Overview. That is Google's account of Google.

In April, two academics published something different. Saharsh Agarwal at the Indian School of Business and Ananya Sen at Carnegie Mellon ran a pre-registered, randomised field experiment through a Chrome extension. They took 1,065 users, randomly assigned them to normal search, search with AI Overviews stripped out, or AI Mode, and watched what happened for a fortnight each. On queries where an AI Overview appeared, outbound organic clicks fell by 39.8% and zero-click searches rose by 34.5%.

Then the part that matters most to a budget holder. Sponsored clicks did not change. Not a little. Not at all. The damage AI Overviews do is to the traffic you were getting for free, not the traffic you were paying for. If your agency has told you that AI is coming for your paid search performance, ask them which measurement that is based on, because the only independent experiment anyone has run says the opposite.

Split Meta's growth and it tells you something useful about your own media plan. Impressions across the Family of Apps rose 14% year on year. Average price per ad rose 12%. So roughly half of Meta's revenue growth is not more advertising being served. It is the same advertising costing more.

That matters because of how these conversations usually go. Your social CPMs are up, your cost per acquisition drifts, and the meeting turns into a question about whether the creative is working or the targeting is right. Sometimes it is. But a 12% rise in the market price of a Meta impression is a market fact, and no amount of creative testing changes it.

Meta also told investors that its new AI ranking systems delivered an 8.3% increase in ad clicks and a 15.7% uplift in conversions on Facebook. Take that at face value if you like, but note that it is Meta measuring Meta, no methodology was disclosed, and there is no independent study that either confirms or challenges it. The company's operating margin, meanwhile, fell from 43% to 31%, which is what the AI build is costing. Someone pays for that eventually, and it tends to show up in the auction.

On 6 July, Sky agreed to buy ITV's media and entertainment business, the free-to-air channels and ITVX, for up to £1.6bn. ITV Studios, the production arm, is excluded and will be spun off separately. The deal needs regulatory approval and completion is expected in the second half of 2027. That is a long way off, but it is already in the room: buyers are reporting that ownership uncertainty is shaping Q4 trading terms right now.

Underneath the corporate news the trading picture is awkward. Total TV grew 0.8% in the first quarter against a full year forecast of 3.7%, while the ad market as a whole grew 9.3%. At the same time Q4 inventory is tightening, because the autumn schedule relaunch lands on top of a full Premier League and a returning Champions League league phase. Scarce audience, soft underlying demand, and a seller with a hard financial reason to protect rate. Read ITV's own announcement and the earn-out is specific: £2 of extra consideration for every £1 by which ITV's total advertising revenue exceeds £1.7bn in the 2027 financial year, capped at £200m and payable in the second half of 2028. ITV is not being paid to sell more airtime. It is being paid to book more revenue, which is not the same thing when the lever available is rate.

Then, on 16 September, the other half of the market moved. Channel 4 Sales became the exclusive UK advertising sales partner for Channel 5 and its portfolio, and for Paramount's MTV, Comedy Central and Nickelodeon, from 2027. Reported as a contract worth more than £300m, and it came off Sky. Put the two together and the shape of UK television sales in 2027 is genuinely unsettled: Sky absorbing ITV's channels while losing Channel 5 and Paramount to Channel 4. Anyone quoting you a three-year deal right now is quoting you a price for a market that will not exist in the same form when the deal matures.

The genuinely new number this quarter is Barb's. In the second quarter, 20.7m UK homes had access to a subscription streaming service, and the ad-supported tiers have quietly become mass reach. Amazon's ad tier is now in 12.1m homes, 40.6% of every home in the country, because Amazon made it the default rather than an option. Netflix's ad tier reaches 7.3m homes and Disney+ 3.6m. Barb is the independent currency, owned jointly by the BBC, ITV, Channel 4, Channel 5, Sky and the IPA, so this is not a platform marking its own homework.

The most recent RAJAR, published on 6 August and covering the twelve weeks to 28 June, is the strongest set of radio numbers ever recorded in this country. 51.1m adults a week, 87% of the population, 1.04bn hours. Digital listening is 76% of all hours and smart speakers alone account for 195m hours a week. Commercial radio passed 40 million listeners and now holds 55.1% of listening against the BBC's 42.6%, a gap of more than nine million people. BBC Radio 2 is still the single biggest station in the country at 12.4m, so the commercial lead is a collective one rather than any one brand overtaking it.

Say the honest thing about the source. RAJAR is jointly owned by the BBC and Radiocentre, the commercial radio trade body. It is the agreed industry currency and the methodology is sound, but a BBC against commercial share comparison is being produced by an organisation both sides own. The reach figures are the part I would take to a board. The share horse race is the part the industry enjoys most.

Widen it out and the MIDAS summer wave, published on 10 August, puts total weekly audio at 56 million adults, 97% of the population aged 15 and over, at an average of 27.5 hours a week. Live and catch-up radio is still 65% of all audio time. Podcasts reach 26% of the population weekly and account for 106m hours. Commercial radio revenue hit a record £747m in 2025, with digital audio revenue up 14.9%.

Outsmart published the second quarter on 3 September: £402.6m, up 12.7% on the same quarter last year, with digital up 14.6% and classic formats up 8.8%. Digital is now 68% of all out of home revenue. Across the first half the medium is up 13.7%. PwC's Andy Lobo noted that digital out of home grew 15.9% in the first half, outpacing the major internet advertising platforms.

Outsmart is the out of home trade body, so the usual warning applies, but this one is worth a qualification in its favour. The figures are compiled by PwC on Outsmart's behalf from member returns, which is a materially more independent process than a channel writing up its own research. I would treat these numbers as sound and the framing around them as sales.

What makes the number interesting is the comparison. Out of home at 12.7% growth, total television at 0.8%, the whole market at 9.3%. Money is not appearing from nowhere. It is moving, and it is moving towards the two things out of home now does well: unduplicated reach that is hard to skip, and digital flexibility that lets a poster behave a bit more like a display buy.