Analysts set the bar at a 6.5 percent decline. WPP cleared it by falling only 4.7 percent. The crowd rose to its feet.

This is what passes for a comeback in adland now: tripping over something lying flat on the ground and being handed a 27 percent standing ovation for the effort.

Shares jumped as much as 30 percent intraday on August 6, closing up roughly 27, the best level since September 2025. Headline operating profit came in at £398 million, beating consensus by more than 13 percent.

That is the entire rally. Not growth. Not a turn. A less-bad contraction, priced like a resurrection.

Cindy Rose, also an ADOTAT reader, six months into the job, wrote in the release that "organic growth remains our North Star." In that same document, organic growth is negative in nine of the ten client sectors WPP bothers to disclose. North America, down 6.0 percent. EMEA, down 4.3. APAC, down 3.8. LATAM, down 1.2. Every region. Negative. Every major business line except Production, which is 5 percent of the book, negative.

You cannot navigate by a star you have never actually seen.

WPP's average headcount fell from 105,958 in the first half of 2025 to 97,490 in the first half of 2026.

That is 8,468 people.

Net sales fell 4.7 percent. Headcount fell 8.4 percent. The 0.2 percentage point margin improvement that the City celebrated with a 27 percent rally is not a story about efficiency. It is not a story about AI. It is not a story about a reimagined operating model or an agentic platform or the trusted growth partner for the world's leading brands.

It is the arithmetic difference between those two numbers. That is all it is.

WPP's own presentation is remarkably honest about this, which I appreciate even as I find it grim. The margin gain came from lower headline severance and cost savings, with staff costs down 5.9 percent to £3,469 million. Translated out of investor-relations English into human: we spent less firing people this year because we already fired them last year.

Which brings us to the part that should terrify anyone building a second-half model.

Total headcount was 98,655 on December 31 and 97,388 on June 30. That is 1,267 people over six months. A 1.3 percent reduction. The 8,468-person year-over-year tailwind is a 2025 event being harvested in 2026 comparisons.

It does not repeat. The lever has been pulled. There is no second 8,468 sitting in a drawer, unless there is, in which case the back half of this year is going to look nothing like the press release.

Every trade outlet on earth reported that adjusted net debt fell £326 million year over year. Almost nobody reported that the leverage ratio got worse.

Adjusted net debt to headline EBITDA rose to 2.18 times, up from 1.98 times. Debt came down. EBITDA came down faster. WPP's own stated target range is 1.5 to 1.75 times.

The company is further outside its own guardrails than it was a year ago, on a day its stock rose 27 percent. Sit with that.

And it thins out further on inspection. Of that £326 million improvement, £125 million is an IFRS 9 accounting amendment adopted January 1. Not cash. Not operations. An accounting change. WPP discloses this clearly in a footnote, to its genuine credit, in the part of the document nobody read.

Real deleveraging is closer to £200 million. WPP also expects over £200 million in disposal proceeds this year.

The debt is going down because the furniture is going out the door.

Reported operating profit: up 18.1 percent.
Reported diluted EPS: down 57.5 percent.

On the headline basis, operating profit down 3.4 percent, headline diluted EPS down 24.5 percent.

That gap is interest on an average adjusted net debt load of £3.3 billion, plus an effective tax rate of 33.5 percent. The majority of what WPP earns at the operating line never reaches a shareholder. It gets eaten in transit by the capital structure.

Meanwhile the interim dividend holds at 7.5 pence against reported half-year EPS of 1.7 pence. That dividend is defensible only if you swallow every single one of WPP's headline adjustments whole, and the gap between reported and headline operating profit this half was £137 million.

And let's be precise about what "maintained" means here. 7.5 pence is not a maintained dividend. It is a dividend that was cut in half in 2025 and is now being held flat at the reduced level. Lying still after falling down the stairs is not the same as standing up.

WPP Open is described in the release as the agentic marketing platform "which enables and connects everything we do."

Technology costs for the half: £319 million, down 6.2 percent.

The AI transformation is being partly funded by spending less on technology. I have read that sentence eleven times and it has not improved.

Gross revenue fell 4.4 percent. Net revenue fell 5.6 percent. Pass-through costs, by our calculation, were essentially flat: £1,628 million against £1,637 million, down roughly half a percent.

The money WPP merely handles is holding steady. The money WPP actually earns for its own work is shrinking faster.

That is a quality-of-revenue problem. It is the slow conversion of an agency into a payment conduit. And it is the one item on this list that does not get fixed by a reorg, because it is not a structural problem. It is a value problem.

ADOTAT covered Martin Sorrell's critique of the WPP Creative restructuring back in February. He told us point blank that this was nothing more than a McKinsey-driven cost exercise dressed up as strategic simplification. He predicted very significant redundancies. He used the word carnage.

Six months later: 8,468 fewer people, margin up two-tenths of a point, and a stock that rallied 27 percent on the arithmetic of exactly that trade.

He was right. It is not close.

Where he was off is the interesting part. He expected a monstrous exceptional charge. The gap between reported and headline operating profit this half was £137 million, real money but nowhere near the tsunami he described.

So one of two things is true. Either WPP is spreading the pain across quarters to avoid printing one genuinely ugly number, or the second half is going to be substantially less pleasant than Thursday's share price implies. Neither is a reason to buy at plus 27.

We also spent March documenting organizational ambiguity inside WPP Media. The overlapping titles. The fiefdoms. The reporting lines pointing in two directions at once. The title that quietly changed after we asked about it.

WPP has now collapsed into a single reporting segment, Global Integrated Agencies, folding the old Specialist Agencies businesses into WPP Media. Fewer disclosure lines. Harder independent comparison. Less for outsiders to check.

And WPP Enterprise Solutions, launched July 1 as the growth engine for enterprise AI transformation, will not get separate revenue disclosure until January 1, 2027.

The new growth story is unmeasurable for eighteen months. Investors are being asked to take it entirely on faith. On Thursday, cheerfully, they did.

On August 3 we argued that WPP's AI push reads as catch-up, not reinvention. These results are the tell. WPP Open is credited with enabling everything and attributed with nothing. Not one pound of revenue. Not one margin point. Not one client win. Anywhere in the release.

Here is where this gets genuinely fascinating, because the most honest read on WPP's competitive position in the AI era came from inside WPP.

Kate Scott-Dawkins is Global President of Business Intelligence at WPP Media. Her job is to synthesize government data, financial results and proprietary spend data into the forecasts this entire industry runs on. Her team tracks more than 200 companies through earnings season. She is very good and she is not a hype merchant.

Asked about AI, she skipped the sales pitch entirely. Large language models, she told us, are "still mostly predictive engines" that guess the next word in a sequence. Her team's actual value, she said, is producing thinking that is "net new to the world entirely," which is precisely the thing those models cannot do. AI is useful to her for summarization and for chewing through more data. It is not useful for the part clients are paying for.

Then she said the quiet part.

The companies at the starting blocks of the AI advertising era, she told us, are the largest sellers of media. She named them: Alphabet. Meta. Amazon. ByteDance. Tencent. She noted that the top 25 sellers of advertising will account for nearly three quarters of the entire industry by the end of this year.

She did not name WPP among the companies at the starting blocks.

She named the companies WPP buys from.

That is not a slip. It is an accurate description of the board, delivered by the person whose entire job is to read the board accurately. And it sits extremely awkwardly next to a results release positioning an agentic platform as the connective tissue of a £9 billion company.

We heard a version of the same instinct from a WPP creative leader we interviewed at Cannes in June, who pushed back hard on the idea that the technology is the product. Data is "not the thing," this executive said, describing work that is "informed by data but not led by data." Creativity, they said, is not going anywhere.

Both positions are defensible. Both are intelligent. Both are also dramatically more modest than what the holding company is selling to the market.

WPP's smartest people describe AI as a tool. WPP's investor deck describes it as a transformation. Those are not the same pitch. The second one is carrying the valuation.

WPP delivered an 8.4 percent headline margin in the first half and is guiding to 12 to 13 percent for the full year.

By our calculation that requires a second-half margin near 16 percent. Roughly double the first half. Producing about two thirds of the year's profit in six months that have not happened yet.

WPP's business is genuinely second-half weighted, so this is not fantasy. But the entire full-year story now rests on a period nobody can verify, while the company simultaneously rebuilds its bonus pool. Staff incentives already jumped to £130 million from £59 million, moving from 1.2 percent of net sales to 2.7 percent. WPP itself concedes second-half margins will fall by up to 200 basis points year over year, partly because of that rebuild.

Cash says the same thing. Adjusted operating cash flow before working capital was £309 million in the first half, down 14.9 percent, against full-year guidance of £800 million to £900 million. The second half has to roughly double the first.

And the client book is not riding to the rescue. WPP's top 25 clients declined 6.3 percent, worse than the group average of 4.7 percent. The biggest relationships are eroding faster than the small ones. CPG is 27 percent of net revenue and down 9.1 percent. Tech and Digital Services is 17 percent and down 9.2 percent. That is 44 percent of the business declining at near double digits.

Now look at the green numbers, because nearly every one of them arrives with an asterisk WPP supplies itself.

China grew 15.6 percent in Q2 "benefitting from timing factors." Spain and Italy grew on "easier comparisons due to prior year one-off factors." WPP Media's improvement was helped because the prior period "included the impact of one-off factors." The second quarter looked better than the first in significant part because the second quarter of 2025 was bad.

That is not momentum. That is a calendar.

None of this makes WPP a corpse. Elevate28 is a real plan. The four-unit structure is a genuine simplification and it was overdue by roughly a decade. The new business list is respectable: Estée Lauder, Henkel, Wendy's are wins by any standard. Stabilization is a legitimate goal and the first half was, by WPP's own framing, on track.

But be exact about what the market rewarded on August 6.

It rewarded a smaller decline than forecast, delivered by a workforce 8,468 people lighter, at a company whose leverage ratio deteriorated, whose earnings per share fell 57.5 percent, whose technology spending declined, and whose newest growth engine will not report a revenue figure until 2027.

The stock went up 27 percent. Nothing in the business grew.

Disclosures: ADOTAT has no financial relationship with WPP plc or any of its operating units, holds no position in WPP securities, and has never sold WPP advertising or sponsorship. This is editorial content. It is not sponsored, and no outside party reviewed, approved or influenced it.

We do not accept paid placement in editorial. We are funded by subscriptions and reader support, and we publish uncomfortable findings about the companies we cover, including the ones whose executives sit down with us.

Figures come from WPP's 2026 interim results release, results presentation and RNS filing of August 6, 2026, plus same-day earnings coverage from Investing.com and Adweek for the share move, consensus comparison, leverage ratio and tax rate. Headcount, staff cost and technology cost figures are from the presentation and RNS. The 8,468 headcount reduction, the pass-through cost comparison and the implied second-half margin are our own calculations from disclosed data; the margin figure is an estimate that rests on second-half assumptions WPP has not published. Interim statements are unaudited.