Nine-month BUMP: 13.2 billion euros, +4.1%, two thirds digital, five historic media at their lowest since 1999. And 67% of advertisers use one lever. My take.
On 18 November, France Pub, IREP and Kantar Media published the BUMP for the first nine months of 2026: 13.237 billion euros of net advertising revenue in France, up 4.1%. Digital, meaning search, social, display and the other online levers, weighs 8.784 billion, up 10.2%, which is 66.3% of the market according to Comarketing-News. The five historic media, television, cinema, radio, press and out-of-home, fall 4.4% to 4.463 billion. "It is the lowest level ever observed since we started measuring the market, that is, 1999", says Virginie Sablé, managing director of France Pub, quoted by The Media Leader. Everyone will read this report as digital's victory. I read something else in it: 69,060 active advertisers, 3% of whom account for 80% of digital investment, and 67% who use a single lever. If you run a scale-up, you are very probably among those 67%. And that is the problem I want to deal with.
French ad market 2026: what the nine-month BUMP says
Here is IREP's table as The Media Leader publishes it, in millions of euros, nine months of 2025 against nine months of 2026. Television: from 2,336 to 2,202, down 5.7%, after -6.2% in the first half. Cinema: from 49 to 54, up 11.5%, the only historic medium that grows. Radio: from 448 to 426, down 4.8%. Press: from 917 to 865, down 5.6%. Out-of-home: from 917 to 915, down 0.3%, with DOOH at 216 million, up 3.6%, now 24% of the medium. Digital: from 7,973 to 8,784, up 10.2%, with a push from the football World Cup. The digital revenue of historic media, the video and audio sold by broadcasters, radio stations and publishers, rises to 635 million, up 8.5%, including 369 million of video at +12.9%. Addressed and unaddressed advertising mail lose another 5% and 5.5%.
Two lines in the report say more than revenue does. Linear TV advertising time falls 15.1%, with thematic channels at -30.1%; radio advertising time falls 9.3%; print advertising pagination loses 8.6%, and 11.6% in regional dailies. In other words, historic media are not just selling cheaper, they are broadcasting noticeably less advertising. When a medium's inventory falls faster than its revenue, either the survivors pay more per second, or the medium has clawed back margin on net prices. In both cases, whoever arrives with a budget and a method has leverage. I will come back to that.
3% of advertisers make 80% of digital: you are in their auction
Kantar Media counts 58,027 advertisers on digital levers, against 21,350 on the five media. Social alone gathers 29,988 advertisers, Amazon's retail media search 5,975, 40% of whom appear on no other measured lever. And 3% of advertisers make 80% of digital investment. Put that next to what I wrote in early November about Meta ad prices, up 14% in a year: when tens of thousands of advertisers concentrate on a handful of platforms, where the largest buy with unlimited budgets, the price of a contact rises for everyone, and it rises most for those with no alternative. A scale-up that built its whole acquisition on Meta and Google is not "digital first". It is a tenant in an auction where it sets neither the price nor the rules, and where its neighbours are Amazon, banks and insurers, the sector investing the most this year at +15% according to Kantar Media.
The 67% single-lever figure is, for me, the real story of this BUMP. It describes a market where most advertisers have never tested a second channel, where the media plan boils down to one platform and one campaign manager, and where nobody knows what the same euro would be worth elsewhere. That is comfortable as long as the CPM holds. It no longer does: +14% at Meta in a year, CPCs rising with AI Mode in search, a retail media market where the challenger brand pays for the leader's incrementality. Staying single-lever in 2027 means accepting inflation you cannot negotiate.
Five media at their lowest since 1999: offline's price is falling, not its value
Here is the other side of the report. The five media at 4.463 billion, their lowest level in twenty-seven years, is 4.463 billion of audience that has not disappeared. The French still watch television, listen to the radio, ride the metro and go to the cinema. What fell is advertiser demand, not available attention. I documented television's third-quarter decline and the rise of programmatic DOOH; the BUMP confirms the trend at market scale. A TV sales house airing 15% fewer seconds, a radio airing 9% fewer, a press losing 8.6% of its pages: these are sellers who need you. Gross rates will not move. Net prices, positions, bonus inventory, reach guarantees, will.
And it works, provided you do not buy offline the way you buy a banner. On Deblock, we delivered more than 30 million impressions by surrounding the metro and geotargeting mobile within 100 metres of stations: a "historic" medium coupled with a digital lever, measured on brand searches and installs, not on a GRP. On Comptoirs de la Bio, isochrone targeting around the stores halved the budget compared with leaflets, 100% tracked: unaddressed print, down 5.5% in the BUMP, replaced by a set-up that costs less and can be measured. DOOH, which 46% of out-of-home advertisers now use against 34% in 2023, is exactly that meeting point: a street medium bought with digital tools. That is what we do with out-of-home and DOOH and with addressable TV and CTV.
What I would do with this BUMP on the table
One, I count my levers: if more than 70% of my budget goes to a single platform, I am in the 67%, and I have a price risk, not a strategy. Two, I take 15 to 20% of the 2027 budget for a second medium, chosen for its current market price, not its fashion: addressable TV, DOOH, audio, cinema if my target is young and urban. Three, I negotiate offline with the BUMP in hand: -15% of airtime in TV, -9% in radio, those are my arguments, not theirs. Four, I measure everything with the same rule, brand searches, direct traffic, sales by area, so the second lever is comparable to the first. Five, I keep the platform for what it does best, conversion, and I stop asking it to create demand at CPMs rising 14% a year.
Who is buying, and who is pulling back: the BUMP by sector
Kantar Media maps gross investment across the five media and digital, up 3.4% overall. At the top: banking and insurance +15%, hygiene and beauty +12%, furniture and decoration +11%, household appliances +10%, real estate and culture-leisure +9%, travel and tourism +7%, energy +6%, education and training +5%. Pulling back: telecoms -7%, automotive and transport -4%, general retail -3%, fashion -2%. A useful reading for a scale-up: if you are a fintech or an insurtech, your direct competitors are the biggest buyers on the market this year, and you will not beat them at auction on the same lever; if you sell fashion or cars, your large competitors are withdrawing, and the space they free up, in television as in outdoor, is there to take at a price they will not see again.
And then there is this line: 10,168 new advertisers in 2026. They weigh only 5% of investment, but Zaïa Ferhaoui of Kantar Media calls them "tomorrow's breeding ground". Kantar cites mobility, with Omoda & Jaecoo or E.Leclerc charging points, and widely varied acquisition strategies, from Bet365 to Laifen. A scale-up that has just raised is, by definition, in that pool. The question is not whether to enter the advertising market: it is to enter through the right door, and not the one where 29,988 advertisers are already crowding.
2027: what Dentsu and WPP Media forecast
The BUMP gives no 2027 forecast; France Pub expects a 2026 communication market of 35.7 billion euros, up 1.3%, with digital at 12.2 billion, up 8.2%, and the five media at 7.6 billion, down 4.1%. For what comes next, CB News compiled the groups' forecasts in October. Dentsu sees the French ad market at +5.7% in 2026, +5.9% in 2027 and +6.1% in 2028; WPP Media about one point lower. Dentsu expects digital at +9% in 2027 then +7.9% in 2028, a television that stabilises in 2027 and returns to +2.4% in 2028, press at -3.1% then +2.7%, out-of-home at +2.2% then +3.1%, cinema at +2% after its +11%. WPP Media puts online retail media excluding search at 1.545 billion euros in 2026, with close to 9% growth a year afterwards, and sees digital reaching 85.3% of media advertising revenue in 2028. My reading: if television hits bottom in 2027 and recovers in 2028, the terms you negotiate this winter are the ones you will not find again. And 2027 is a presidential election year, which radio sales houses, according to The Media Leader, already describe as a weight on every strategy: attention will be expensive in the spring, and discounted elsewhere in the year.
What I tell founders
November's BUMP tells the story of a market that is concentrating: two thirds of the money on digital, 80% of that digital in the hands of 3% of advertisers, and a majority of advertisers with a single lever who suffer the inflation caused by the others. Meanwhile, historic media are at their lowest since 1999, with less advertising on air and sellers who need to sign. I am not telling you to leave Meta or Google. I am telling you that a scale-up's 2027 media plan cannot rest on one platform whose price rises 14% a year while the inventory across the street is negotiated downwards. The media mix simulator gives you a first idea of what a second lever changes in your CAC. And if you want to know precisely where your budget is exposed to this concentration, and what share to move out, that is the work I do during a free audit.