
Nation Media Group’s 2025 annual report counts 63.6 million monthly digital users. The same report records revenue of KSh 6.04 billion, down 3.1%, and a third straight annual loss. Divide one figure by the other and the group earns about KSh 95 a year for each monthly user, and that counts every shilling from print and broadcast as well. The gap between audience and income is the attention recession as it reaches East Africa: more people within reach than ever, and less money for each person reached.
Key Takeaways
- Nation Media Group’s revenue fell from a KSh 13.4 billion peak in 2013 to 2014 to KSh 6.04 billion in 2025, while its digital audience grew past 63 million monthly users. Reach and revenue have come apart.
- In the first half of 2026 the group lost KSh 357.2 million on revenue down 4.8%, a loss larger than its whole loss for 2025.
- Pew Research Center found that U.S. Google users clicked a traditional search result on 8% of visits when an AI summary appeared, against 15% on pages lacking one. Google disputes the method.
- In the Reuters Institute’s 2025 survey, people rated platforms above news media for local buying, selling and services; news media kept the lead on local politics, local news and notices.
- The Attention Ledger sorts each revenue line by the attention it rests on: commodity information, rented reach, owned relationships or trust. The first two columns are shrinking; the last two are where a local media business can still set its price.
- The phrase “attention recession” was coined for saturated Western markets. In Kenya, with internet penetration at 35%, total attention is still growing; the recession sits in the share that pays local publishers.
Where the phrase “attention recession” came from
The analysts at MIDiA Research used the term through the pandemic years. Their argument was simple arithmetic: the supply of content kept growing after the lockdown surge in screen time, while the hours people could give it stopped growing. When Meta reported its first ever fall in daily active users, from 1.930 billion to 1.929 billion, MIDiA called it the moment the “attention recession” showed its teeth, and noted that about $200 billion came off Facebook’s market value after the report.
For a media business the definition is practical: attention becomes a fixed plot with more tenants every season, and every new streaming app, TikTok creator and AI assistant takes a slice of the same evening.
East African publishers met it through four channels at once. Audiences moved to social video, advertisers cut budgets, and AI search began answering questions before anyone clicked through. In Kenya, some of the largest customers, government entities among them, also turned out to be the slowest payers.
The Nairobi ledger: audience up, revenue down
Nation Media Group is the best documented case in the region, because it is listed in Nairobi, Kampala, Dar es Salaam and Kigali and must publish its numbers.
Full-year turnover fell from KSh 7.61 billion in 2021 to KSh 6.04 billion in 2025, a drop of 20.7%, and the group posted lower revenue in every year after 2021. Against the KSh 13.4 billion peak of 2013 to 2014, revenue is down by more than half. The net margin moved from a positive 6.5% in 2021 to negative 5.1% in 2025. Losses ran KSh 205.7 million in 2023, KSh 254.4 million in 2024 and KSh 308.6 million in 2025. Then the first half of 2026 brought a loss of KSh 357.2 million on revenue of KSh 2.85 billion. Cash and short-term investments stood at KSh 1.89 billion in June 2026, against KSh 3.80 billion in June 2022.
The growth lines are real. Digital revenue and broadcasting revenue each rose 5% in 2025. Paywall subscriptions grew 13% and ePaper subscriptions 10%. The trouble is scale: print is falling faster than the new lines are rising. One honest wrinkle on the audience figure: the annual report gives 63.6 million monthly digital users and the abridged results give 64.7 million. This essay uses the lower number.
The agency side of the market tells the same story from the buyer’s chair. WPP Scangroup, the Nairobi-listed advertising group, reported FY2025 revenue down 16.4% and gross profit down 27%, which Mwango Capital’s review attributed to sustained client budget cuts. When the people who buy space are shrinking, the people who sell space shrink with them.
Two more details belong in a founder’s risk register. The Reuters Institute’s Kenya report records that Safaricom withdrew advertising from Nation after critical reporting on the company. And in its 2026 half-year results Nation booked higher provisions for uncollected debts, particularly delayed payments from government entities. Advertising concentrated in a few large accounts is an editorial risk and a cash-flow risk at the same time.
Search is sending fewer readers
For a decade, search engines delivered local publishers a steady stream of readers. AI summaries are thinning it. Pew Research Center analyzed the browsing of 900 U.S. adults and found that 58% ran at least one Google search in March 2025 that produced an AI summary. On pages with a summary, users clicked a traditional result in 8% of visits; on plain results pages, 15%. They clicked a link inside the summary itself in about 1% of visits.
Across the markets in the Reuters Institute’s 2025 survey, 7% of people used an AI chatbot for news each week, and 15% of under-25s did.
The strongest case against all this comes from Google itself. In a statement to The Register, the company called Pew’s method flawed and its query set skewed, and said: “We consistently direct billions of clicks to websites daily and have not observed significant drops in aggregate web traffic as is being suggested.” I concede the point on its own terms. Google sees aggregate traffic and Pew sees a U.S. panel; Kenyan, Ugandan and Tanzanian readers sit outside both. In a region still adding internet users every year, the total number of clicks may well keep rising.
The local publisher, though, gets paid on its own clicks. A reader who gets the answer from a summary has spent her attention on Google’s page, and whatever that visit earns stays with Google. Aggregate growth and local decline can both be true in the same year, and the Nation ledger suggests they are.
The Attention Ledger
The Attention Ledger is a way to read your own revenue by asking one question of each line: what kind of attention does this shilling depend on? It has four columns.
Commodity information. Bus times, weather, listings, prices, job adverts, buy-and-sell. The Reuters Institute’s 2025 chapter on local news found that, averaged across markets, people now rate platforms above news media for local buying and selling (net minus 23 points), local services (minus 11) and local activities (minus 6). Richard Fletcher, who wrote the chapter, argues that platforms win wherever the information is the same everywhere and convenience is the only difference. Revenue in this column is leaving, and good journalism will struggle to call it back.
Rented reach. Views on Facebook, YouTube and TikTok, sold to advertisers by the thousand or paid out as platform revenue share. The audience is real, but the algorithm owns the tap. The Reuters Institute counts six online networks reaching more than 10% of people weekly with news, against two a decade ago, and reports that its Kenyan respondents now prefer watching news to reading it. Video reach is necessary, and it is the column most exposed to a policy change written in California.
Owned relationships. The WhatsApp channel, the SMS list, the app, the paywall subscriber, the reader who opens your newsletter on Monday morning. Nation’s 13% paywall growth sits here. These are people you can reach on your own terms, and the economics of repeat purchase apply to them the way they apply to any retailer.
Trust. The reason people come to you specifically when the stakes are high. The same Reuters chapter found news media still preferred for local politics (plus 16), local news (plus 13) and notices such as deaths and marriages (plus 10). In Kenya, overall trust in news stood at 65% in 2025, and Citizen, NTV and the Daily Nation were each trusted by around 90% of the sample. Over the same year, weekly use of Citizen TV in that sample fell from 76% to 68% and of the Daily Nation from 59% to 51%. Reach fell while trust held. Trust appears on nobody’s balance sheet, yet it is the asset the attention recession leaves standing.
Running the ledger takes an afternoon. List last quarter’s revenue by source, mark each line with its column, and pick one line to move rightward this quarter: a notices column turned into a verified paid announcements service, an advertiser relationship rebuilt as a sponsored briefing for a defined trusted audience, a weekly report sold to the NGOs and institutions that already read you for free. Each of those trades a commodity or rented line for an owned or trusted one.
Where the Western frame fails in East Africa
The attention recession was named in markets where nearly everyone was already online and the hours in a day had run out. Kenya’s internet penetration was 35% in the Reuters Institute’s 2025 figures. Here the pool of attention is still filling. So the East African recession is distributional: more total attention, flowing to fewer and larger platforms, with less of it reaching the local publisher’s account.
The evidence also has holes, and a founder should know where they are. The Reuters Institute’s Kenyan data represents younger, English-speaking online users aged 18 to 50, and the authors say plainly that it is unrepresentative of the nation. Vernacular radio audiences and offline households sit outside the survey, and those are exactly the audiences many district stations serve. Uganda and Tanzania sit outside it altogether: in Africa the 2025 report covers Kenya, Morocco, Nigeria and South Africa only. A Kampala or Mwanza founder is working from Kenyan proxies and should hold these conclusions more loosely.
One finding may transfer with extra force. Fletcher’s chapter notes that dependence on platforms for local information tends to run highest where the newspaper industry was historically weak. Markets with thinner print histories may therefore see platforms take over local information faster. That is my inference, offered as one: the commodity column may empty sooner in a smaller market.
The trust column has a regional opening. In the 2025 survey, 59% of Kenyan respondents named online influencers and personalities as a major source of false or misleading information, and the report places Kenya and Nigeria where this concern runs highest. A public worried about what is true will pay, in some form, for a source it can check.
A good name on the balance sheet
The attention economy frame deserves its due. It treats attention as the scarce resource and asks who can win the most of it at the lowest cost, and that logic has handed small Kampala founders cheap distribution that earlier publishers could only dream of buying. On its own terms, though, it lacks a line for the thing the Kenyan numbers keep pointing to. Reach fell; trust held at 65% and brand trust near 90%. The frame measures attention captured this week; a reputation built over decades and lost in one bad story sits outside its arithmetic. Proverbs 22:1 prices it plainly: “A good name is more desirable than great riches; to be esteemed is better than silver or gold” (NIV). For those of us building media businesses, that verse describes the asset the attention recession leaves in our hands. Truthful reporting is stewardship of a neighbor’s trust, and a business run that way owns the one column of the ledger that grows in value as attention gets scarcer.
Frequently Asked Questions
What is the attention recession?
The attention recession is the point at which the supply of content keeps growing while the time people can give it stops growing. MIDiA Research used the term as the pandemic surge in screen time faded. For media businesses, each new competitor takes a slice of a fixed pie, so reach gets cheaper to find and harder to sell.
Is AI search cutting traffic to East African news sites?
The best public measurement is American. Pew found U.S. Google users clicked a result on 8% of visits with an AI summary, against 15% on plain pages. Google disputes the method and reports stable aggregate traffic. My research found only American measurements, so treat the direction as likely and the East African size as unmeasured.
Why is Nation Media losing money if its digital audience is growing?
Print revenue is falling faster than digital and broadcast revenue are rising. In 2025 digital and broadcasting each grew 5% and paywall subscriptions grew 13%, but total revenue still fell 3.1% to KSh 6.04 billion. Agency budget cuts and delayed government payments added pressure. The audience grew; the money earned per reader shrank.
What can a small local media business sell besides advertising reach?
Sell what the Attention Ledger places in the owned and trusted columns: paid verified notices, sponsored briefings to a defined audience, reports for NGOs and institutions, memberships and events. The Reuters Institute found local news media still preferred over platforms for local politics, local news and notices, so build products on those strengths.
Does the attention recession apply in Uganda and Tanzania?
Partly, and the evidence is thin. The Reuters Institute’s 2025 survey covers Kenya and leaves out Uganda and Tanzania, and its Kenyan sample skews young, urban and English-speaking. Total attention is still growing as more people come online. The pressure shows up in which platforms capture that attention, and how little of it reaches local publishers.
Next Step
Before Friday, pull last quarter’s revenue, sort every line into the four columns of the Attention Ledger, and choose the one line you will move toward owned or trusted attention this quarter. If you want an outside read on whether your team and systems can carry that shift, take the AVODA Business Capability Assessment. It will show you where your operating gaps sit before you spend a quarter building a product your back office cannot support.
Related Reading
- GTM in the AI Era: AEO Is the New SEO for Founders
- Retention Is the Cheap Growth Engine You’re Ignoring
- The B2I Playbook: Selling to NGOs and Institutions
- Cash Flow When Everyone Pays Late: The 90-Day Economy
Sources and Evidence
- Nation Media Group Limited (NMG.ke) 2025 Annual Report: FY2025 revenue of KSh 6,039.4 million (down 3.1%), 63.6 million monthly digital users, paywall subscriptions up 13%, ePaper up 10%, digital and broadcasting revenue each up 5%.
- Nation Media Group Limited (NMG.ke) 2025 Abridged Report: loss after tax of KSh 308.6 million; digital user base of 64.7 million (the figure that differs from the annual report).
- CEO East Africa: Nation Media Group’s H1 net loss widens to Kshs 357.2 million: H1 2026 loss and revenue; turnover from KSh 7.61 billion (2021) to KSh 6.04 billion (2025), down 20.7%; losses for 2023 to 2025; net margin from 6.5% to negative 5.1%.
- Kenyan Wall Street: Nation Media Group’s revenue falls to 21-year low: H1 2026 provisions for delayed government payments; cash and short-term investments of KSh 1.89 billion against KSh 3.80 billion in June 2022.
- Mwango Capital: Nation Media posts third straight loss: KSh 13.4 billion revenue peak in 2013 to 2014; WPP Scangroup revenue down 16.4% and gross profit down 27%.
- Reuters Institute Digital News Report 2025: Kenya: 35% internet penetration; trust at 65%; brand trust near 90%; Citizen TV and Daily Nation weekly reach declines; Safaricom advertising withdrawal; sample limits.
- Reuters Institute DNR 2025: Local news, how publishers can still provide value in a platform world: net preference scores for platforms versus news media by local information type; path dependency on historic newspaper strength.
- Reuters Institute DNR 2025: Executive summary and key findings: 7% weekly use of AI chatbots for news (15% of under-25s); six networks above 10% weekly news reach; Kenyan preference for watching news; 59% Kenyan concern about influencers.
- Reuters Institute Digital News Report 2025 (full PDF): African markets covered (Kenya, Morocco, Nigeria, South Africa); Kenyan sample representative of younger English speakers aged 18 to 50.
- Pew Research Center: Google users are less likely to click on links when an AI summary appears: 900 U.S. adults; 58% saw at least one AI summary in March 2025; fewer clicks on pages with summaries.
- The Register: Google AI Overviews and the Pew study, with Google’s statement: 8% versus 15% click rates, 1% clicks on summary links, and Google’s dispute of the method.
- MIDiA Research: The attention recession just showed its teeth to Facebook: origin of the term; Meta daily active users from 1.930 billion to 1.929 billion; $200 billion market value loss.
