Revenue Editors Picks

Britain’s ad boom exposes publishers’ diversification problem

Ivan Massow 7 Nov 2026 · 5 min read

A red-jacketed advertising buyer pushes an oversized gold shopping trolley through a narrow doorway, squeezing a blue-jacketed publisher holding a newspaper against the doorframe. The caption reads: There’s room for everyone.

Cartoon · AI-generated

Display advertising grew, yet digital publishers’ total revenue fell. The AOP and Deloitte release on 4 November puts participating UK publishers’ April-to-June revenue at £155.04 million, down 1.49% on a year earlier, despite 4.50% growth in display income and rising subscriptions.

The warning is about the mix, not the death of advertising. Publishers with a stronger core ad business can still lose ground elsewhere. Advertisers and retailers have more ways to trade on audience attention; publishers bear the cost of building products that may depend on the same distribution or buying systems. The decision is which new income changes that exposure, rather than merely adding another format.

Andy Cowen, Deloitte’s lead partner for telecoms, media and entertainment, described the picture in the release as “increasingly one of divergence rather than uniform growth or decline”. Neither an industry recovery story nor an indiscriminate retreat from advertising captures that.

The wider market makes the contrast sharper. AA/WARC’s October release puts UK advertising expenditure up 11.0% to £12.3 billion in Q2. Retail media grew 31.2% and social media 22.2%. Yet regional newsbrands’ online advertising fell 13.2%, while the national newsbrand online line edged up 0.4%. The same quarter offered very different markets to sell into.

Advertisement · Noah

“Where that money is going is changing too,” Elizabeth Lane, head of insight at IAB UK, said in its September Digital Adspend announcement. Its separate first-half study with mediasense recorded 18% growth in video spending to £5.1 billion, with a third going to TV+, internet-delivered video on large television screens.

Those video figures are not an estimate of the budget available to a newspaper’s clips. Period, inventory and coverage differ from the publisher survey. A growing format can contain markets in which a publisher has little inventory or bargaining power. Moving into video changes the output. It does not, by itself, establish access to the spending.

When the new lines weaken

Miscellaneous revenue in the AOP panel fell £3.93 million, exceeding the £2.35 million net decline in total income. Off-platform revenue fell 16.11% and video 6.00%. The release tentatively links miscellaneous weakness to AI, without quantifying that connection. Calling the fall an AI loss would go beyond the evidence.

AOP also suggests that adding video to audio-only podcasts may explain some audio weakness. Revenue moving between columns can look like a change in the business. It does not establish that the lost income has all reappeared.

Consider a deliberately hypothetical £100 revenue portfolio. Display contributes £60 and other activities £40. If display grows 5%, it reaches £63. If the remainder falls 15%, it reaches £34, leaving total revenue at £97. The rates and amounts are invented assumptions. The example shows why a healthy core and a shrinking business can coexist; it says nothing about the cause of either movement.

This is a participant benchmark, not a census. The current roster is undisclosed. Dividing a total that includes reader income by national advertising expenditure would not produce a meaningful publisher market share.

Even the advertising labels need care. AA/WARC’s “other online display” explicitly excludes news and magazine brands, as well as several other channels. A decline in that line cannot be substituted for publisher advertising. Commercial decisions require the relevant market, not simply a familiar-sounding category.

Reach shows the importance of following the definitions into company accounts. Its subsequent third-quarter update reported indirect revenue down 29.7%, against a 5.5% fall in direct revenue. It also said improving off-platform growth, including AI licensing, partly mitigated the impact of lower on-site audiences. Those are Reach’s categories and explanations, not the AOP panel’s.

A different claim on the budget

Retail media, advertising sold through retailers’ environments or using their customer data, offers one route towards expanding budgets. Future’s partnership with Ocado Ads and Permutive is a concrete UK example. Announced in July 2025, it was designed to combine Future’s audience information with Ocado’s shopper data through Future’s Aperture advertising product.

The logic is intelligible. A specialist publisher knows something about what a reader is considering; a retailer has information about purchases. Together they may give an advertiser a more useful audience than either can supply alone. That could help a publisher compete for a different part of the marketing budget, rather than sell more impressions at the same price.

But the announcement describes a proposition, not the publisher’s fees or net gain. Future’s broader Aperture page reports case-study improvements, including 2.6 times the return on advertising spend for an unnamed big-box retailer. It does not identify that advertiser as Ocado or demonstrate a controlled increase in sales from this partnership. Reaching people who buy is not the same as causing the purchase.

Nor does a data partnership automatically remove dependence. It can replace some reliance on search traffic with reliance on a retailer, a technology supplier or a particular advertiser’s measurement system. That may be a good trade. It has to be assessed as a commercial relationship, with its own costs and bargaining position, rather than celebrated as diversification simply because the revenue receives a new label.

Even a direct audience relationship leaves another dependency intact. WebKit’s 19 October explanation confirms that Safari blocks connections to identity-resolution services, which link identifiers such as email addresses across websites. IAB Tech Lab proposes moving more advertising operations onto publisher-controlled servers. Its own explanation says production evidence is still being developed. The implication is narrower than a predicted revenue collapse: control of the audience does not confer control of the technology through which advertising reaches it.

Offsets are possible

People Inc.’s latest US results provide a counterexample to the idea that falling visits dictate the same outcome everywhere. In Q3, digital revenue grew 2% even as advertising fell 8%. Licensing and other income rose 49%, helped by Apple News+, syndication partners and the Meta content agreement. Affiliate commerce grew 4%, which the company attributed primarily to higher transaction volumes.

This is a different country, quarter and business mix. It cannot explain the AOP aggregate or supply a UK benchmark. It does show that alternative income can offset advertising weakness. It also shows why the source of that income matters: licensing to large platforms can support revenue while leaving significant negotiating power with those platforms.

The useful distinction is between an extra product and an additional reason for a customer to pay. A video sold to the same advertiser, through the same platform and against the same audience, may broaden the product range while leaving much of the original exposure intact. A licensing contract or a purchase-data partnership can change the proposition more substantially, but introduces terms and counterparties that need to earn their place.

Britain’s advertising growth is real. It is not evenly accessible, and the publisher survey shows that growth in one established line need not rescue the whole portfolio. The strongest commercial response is to identify the budget a publisher can credibly win and the value it can demonstrate there. More revenue headings on a management report are a poor substitute for that evidence.

Research note: Noah, a research and forecasting service, was used to examine publishing coverage and test follow-up questions. Material claims were checked against the primary sources below. No numerical forecasts were available for publication. Additional definitions and limitations are in the linked research notes.

Sources

Source: Tomorrow’s Publisher