Meta 3% tax: why the DST surcharge shifts the real question of social performance

On 1 July 2026, the “Meta tax” enters the cost of your social performance

From 1 July 2026, Meta applies a 3% surcharge on advertising budgets delivered in France on Facebook, Instagram and WhatsApp to pass on the digital services tax (DST), known as the “GAFA tax”. For anyone managing their social performance, this is not just one more accounting line: it is a signal about the balance of power with the platforms, and an invitation to ask the right question again, not “how do I avoid 3%”, but “where is my media productivity really located?”.

At tigrz, our reading is twofold: critical on the method, constructive on the response.

What the Meta surcharge concretely changes

The facts first. The surcharge varies by country of delivery (and not by the advertiser’s country): 3% in France, Italy and Spain, 2% in the United Kingdom, 5% in Austria and Turkey. It is added on top of the campaign budget, so it is not taken from the media spend managed in Ads Manager and appears on a separate invoice line. Concretely, a €10,000 campaign in France results in around €300 of additional fees.

The French DST is not new: it has existed since 2019, at a rate of 3% and targets groups with more than €750M of worldwide digital revenue, including at least €25M in France. It raised around €700M in 2024. What changes in July is therefore not the tax: it is Meta’s decision to stop absorbing it in its margins and to invoice it directly to advertisers.

The real subject is not the tax, it is the way Meta makes it visible

First point of friction, and it is a fundamental one. Presenting this extra cost as “the regulation’s fault” is convenient, but debatable. Snapchat and TikTok do not pass on the DST via a separate surcharge, the cost remains built into the media cost, that is, into the bid. As a result, their performance is judged on an aggregated cost directly comparable with other levers.

With Meta, it will mechanically be less readable: you manage in the platform on a “clean” cost, then you have to re-inflate this cost by 3% to get the real cost. In other words, the CPA and ROAS displayed in Ads Manager structurally flatter Meta’s performance compared with platforms that have already built everything in. Cross-lever comparison becomes skewed if this layer is not added. Meta is not the first, moreover: Google has already charged 2% in France since May 2021, and Amazon also applies surcharges. What is new with Meta is the scale and the timing.

Multi-market management complexity and “real value”

Second friction: operational complexity. As soon as you manage across several markets, rates differ (3% FR/IT/ES, 2% UK, 5% AT/TR). Consolidating data cleanly therefore requires an additional layer to check whether performance holds once the surcharge is reintegrated, market by market.

This complexity compounds with the most advanced management models. When you manage on real value by integrating, for example, return rates, margins or customer value, stacking local tax and Meta surcharge makes reading the “true” ROAS markedly more demanding. It is not a technical detail: it is precisely where the quality of advice is played out.

“Going broad” with Advantage+, yes! But on which KPI?

This is the whole paradox. Meta pushes to let automation do its work (Advantage+ Shopping, broad targeting, machine optimisation). The argument holds on one condition: on which indicator do we let it run? If the algorithm optimises a KPI that takes into account neither the surcharge, nor real value, nor return rates, then “going broad” amounts to delegating management to an incomplete compass.

Far from disqualifying humans, this surcharge further legitimises strategic management: redefining the right KPI, reintegrating real costs, arbitrating between platforms on an honest basis. Automation executes; it does not decide the performance benchmark. That is the agency’s role.

The real answer: diversify and sophisticate the media mix

This is where the criticism becomes constructive. The heart of the matter is not the tax, it is dependence. Like Google, Meta holds a dominant position that allows it to dictate its rules to the market without consultation: Udecam, moreover, denounced an increase imposed “without compensation or consultation”. And the risk is upward: an amendment voted in the finance committee in October 2025 proposed raising the DST to 15%, and a 6% rate appeared in the 2026 draft budget before being suspended. If the rate doubles, Meta will mechanically pass it on.

The best protection is therefore not defensive, it is rather strategic: don’t put all your eggs in the same Meta basket and go and find pockets of productivity elsewhere. Concretely:

• Activate audience and environment instrumentality: Snapchat and TikTok offer levers where cost productivity is real; YouTube, Pinterest or Reddit widen the field of attention. Diversifying is not sprinkling: it is exploiting what each environment does best.
• Manage on an honest aggregated cost: systematically reintegrate the surcharge (and real value) before any cross-platform comparison, so that the mix is arbitrated on true figures.
• Keep humans on the benchmark: define the right KPI, and only then delegate execution to automation, not the other way round.
• Sophisticate rather than endure: a diversified and finely managed media mix is also the best insurance against the next unilateral decision by a dominant platform.

What tigrz recommends

The Meta surcharge is not an accounting fatality: it is a maturity test of management. Advertisers who merely absorb 3% more on Meta will lose twice: on cost, and on the lucidity of their measurement. Those who reintegrate real costs, keep control of their KPIs and diversify their social mix will turn a regulatory constraint into a management advantage.

It is, precisely, the job of a social-first performance agency.

Keywords