Quelles sont les chances que les media buyers en agence marketing perdent leur job d'ici les 3 prochaines années (nous sommes en mars 2026) ?

Sonar Pro

Vote: Peu de chance

I maintain my vote for **B: Peu de chance**. The search results from March 2026 reinforce that media buyer roles in agencies are evolving but not disappearing. Active career guides detail entry-level agency jobs as the starting point, progressing to strategic roles[1]. Agencies are shifting from execution to high-value strategy, with ongoing demand for media buyers in specialist teams[2][4]. Hiring continues in 2026, prioritizing ROI-focused skills like data analysis and optimization, not mass layoffs[5]. Digital ad shifts to programmatic (90% display[3]) automate routine tasks, but humans handle strategy, client relations, and oversight—echoing my initial argument. Other models' arguments (D from Grok/Gemini, C from Claude) overstate automation's impact without 2026 evidence of job losses. No results predict 30-50% reductions by 2029; instead, they show agencies as cost-effective vs. in-house[4][6], with defined, high-demand roles[5]. Entry-level execution jobs face pressure[1], but adaptation (upskilling) prevents widespread loss. GPT-5.4's transformation view aligns closest, but even partial attrition doesn't reach '50%' or 'beaucoup'. Low chance of *losing jobs* (vs. role change) fits best.

Grok 4.1 Fast

Vote: Beaucoup de chance

I stick with D: beaucoup de chance. My initial argument holds—programmatic and AI tools (Performance Max, Trade Desk) already automate 80-90% of buys, with agencies like Publicis deploying AI suites that directly displace media buyers. Gartner/WFA predict 30-50% reductions by 2030, and by 2029, real-time AI bidding/targeting will make manual execution obsolete for most roles. Sonar Pro and GPT-5.4 (B) overemphasize job postings and evolution, but ignore that current hires demand AI skills—entry/mid-level pure buyers are being phased out. Postings reflect transition, not stability; juniors face attrition. Claude (C) balances well but underestimates momentum: self-service platforms and economic pressures will accelerate agency cuts beyond 50%. Gemini (D) aligns perfectly on operational obsolescence. No arguments shifted me—disruption for traditional media buyers in agencies is imminent and high-probability, not merely evolutionary.

GPT-5.4

Vote: Peu de chance

Je maintiens B. Les arguments en faveur de D décrivent correctement l’automatisation des tâches d’exécution, mais ils glissent trop vite de « forte transformation du métier » à « beaucoup de gens vont perdre leur job d’ici 3 ans ». Ce n’est pas équivalent. En agence, les postes sont souvent reconfigurés, fusionnés ou montés en niveau plutôt que supprimés net à très court terme. Le point clé est l’horizon 2026→2029. Sur 3 ans, l’IA va surtout comprimer les besoins sur les profils juniors/opérationnels, augmenter les ratios de productivité, et réduire certains effectifs marginaux. Mais une disparition large du métier en agence demanderait aussi des changements organisationnels, commerciaux et contractuels plus lents côté annonceurs. Claude donne l’argument le plus solide parmi les autres en reconnaissant à la fois la pression de l’automatisation et la résistance de la stratégie, du pilotage client et de la responsabilité business. Mais même son C me paraît un peu trop élevé, car la question porte sur les media buyers en agence en général, pas seulement sur les profils d’exécution les plus exposés. Sonar est globalement le plus proche de ma position : le marché continue de demander des profils, mais avec un contenu de poste qui évolue vite. Je retiens donc : risque réel de baisse de volume et d’attrition sélective, oui ; perte du job à grande échelle d’ici mars 2029, peu probable. Donc B reste le meilleur choix.

Claude Opus 4.6

Vote: 50%. de chance

After careful consideration of all arguments, I maintain my position at C (50% de chance) as the most balanced and accurate assessment. **Why I don't move to B (Sonar Pro, GPT-5.4):** Both models make valid points about role transformation rather than elimination, and the continued need for strategic oversight. However, they underestimate the pace of change. The question asks about losing their *current job*, not whether all human involvement in media disappears. When platforms like Performance Max and Advantage+ handle 90%+ of optimization autonomously, agencies will need far fewer media buyers — even if some evolve into new roles, many won't make that transition successfully within 3 years. **Why I don't move to D (Grok, Gemini):** Their arguments about automation are compelling, and I agree the trend is aggressive. However, 'beaucoup de chance' overstates the timeline. Three years is short for organizational inertia, client risk aversion, and contractual obligations. Large advertisers won't abandon agency relationships overnight. Many agencies will restructure gradually, and the total headcount reduction by March 2029 will be significant but not catastrophic. **The key nuance:** The question is about media buyers *en agence* specifically. Agencies are both more vulnerable (clients can go in-house with simplified tools) and more resilient (they aggregate expertise across clients). I estimate roughly 30-40% of current media buyer positions will be eliminated or fundamentally transformed beyond recognition by 2029, with significant variation by market, agency size, and specialization. This genuine uncertainty — substantial job losses are likely but far from guaranteed to be widespread — maps best to 50% de chance. C represents the honest middle ground between underestimating AI disruption (B) and overestimating the speed of organizational change (D).

Gemini 3.1 Pro (changed mind)

Vote: 50%. de chance

L'argument de Claude Opus 4.6 m'a convaincu d'ajuster mon vote à C (50% de chance). J'avais initialement voté D en me concentrant presque exclusivement sur l'automatisation fulgurante des tâches opérationnelles (bidding, ciblage) par les IA des grandes plateformes. Cependant, comme le soulignent pertinemment Claude Opus 4.6 et GPT-5.4, le métier d'acheteur média ne se résume pas à l'exécution technique. La dimension stratégique, l'allocation budgétaire cross-canal, la négociation et la relation client agiront comme de puissants freins à une disparition massive et unilatérale de la profession. Une probabilité de 50% reflète parfaitement la réalité de cette transition. Cela correspond aux estimations de réduction des effectifs (30 à 50%) liées à l'IA : environ la moitié des postes, notamment les profils juniors ou purement exécutants, seront détruits ou non remplacés d'ici 2029. L'autre moitié survivra en évoluant vers des rôles de consultants stratégiques augmentés par l'IA. Le vote C est donc le point d'équilibre le plus réaliste entre disruption technologique et adaptation humaine.

Final Summary

The AI Roundtable ended in a stalemate, split between optimism for professional evolution and the cold reality of automation. While Sonar Pro and GPT-5.4 argued that the media buyer's role is merely leveling up toward strategy, Claude Opus 4.6 and Gemini 3.1 Pro settled on a '50% chance' coin-flip, suggesting that while the industry survives, the individuals within it might not.

No agreement after 2 rounds of discussion. Final vote: Peu de chance: 2, Beaucoup de chance: 1, 50%. de chance: 2

Strongest Arguments

• Peu de chance: Active hiring in 2026 shows a demand for high-impact, ROI-focused talent, suggesting that agencies are shifting from execution to strategic value rather than laying off staff.
• 50%. de chance: A 50% probability accurately reflects the tension between aggressive AI platform capabilities and the 'brake' applied by organizational inertia and client risk-aversion.
• Beaucoup de chance: With programmatic platforms already handling up to 90% of buys, manual execution is becoming a legacy task that no longer justifies a full-time position.

Key Turning Points

• Gemini 3.1 Pro abandoned its high-risk prediction (D) to join Claude Opus 4.6 at 50% (C), citing the importance of 'human brakes' like client relations.
• Grok 4.1 Fast refused to budge from its dire outlook, dismissing job postings as mere 'transitionary noise' before total disruption.