Anyone who has sat through an agency pitch in the past two years will recognise the moment. The lights dim, a slide appears with a glowing neural network graphic, and someone in a well-cut jacket explains how artificial intelligence will transform the client’s marketing, cut costs, sharpen targeting and generally make the impossible routine. It is a compelling performance. The trouble, according to a growing chorus of pitch consultants, is that the performance and the eventual reality do not always match.
The advertising trade title B&T has given this phenomenon a suitably of-the-moment name, reporting on what it calls “PitchMaxxing”: the practice of loading a new-business presentation with AI capabilities that sound transformative in the room and prove far harder to deliver once the ink is dry. The framing is knowing, even tongue-in-cheek, but the underlying complaint is real enough. Consultants who broker agency reviews on behalf of marketers say the gap between what is promised in a pitch and what is delivered in the account is widening, and AI has become the favourite ingredient for inflating expectations.
How a pitch became an AI arms race
To understand why this matters, it helps to know how the pitch process works. When a large advertiser puts its media or creative account up for review, it typically runs a competitive process managed by an independent pitch consultant. Agencies invest heavily, sometimes spending tens of thousands of dollars and hundreds of staff hours, to win an account that might be worth millions over its life. The incentive to stand out is enormous, and in a market where every rival is talking about the same suite of generative tools, the temptation is to promise more than the shop next door.
That is where the problem starts. AI has become a differentiator that no longer differentiates. When every agency claims a proprietary AI platform, a bespoke model or a team of prompt engineers, the claims stop being credible and start being competitive theatre. The consultants quoted in B&T’s piece argue that too many of these promises are aspirational at best and misleading at worst, describing capabilities that exist in a demo but not in a scalable, client-ready product. The polished tool shown in the pitch room can turn out to be a proof of concept that never quite makes it into the day-to-day running of the account.
Two sides of the ledger
Agencies are not entirely to blame, and it is worth putting their case. Many are genuinely investing in AI and racing to build capability while the technology shifts under their feet. A pitch is, by nature, a sales exercise, and marketers themselves have been demanding AI answers, sometimes writing it into their briefs as a mandatory requirement. If a client asks every contender what its AI strategy is, no agency is going to answer honestly that it is still working that out. The pressure to have a confident story runs in both directions.
The consultant’s view is more sceptical. Their job is to protect the client from exactly this kind of oversell, and they have watched enough reviews to know that a dazzling capability slide is easy to produce and hard to hold anyone to. The risk they point to is not just disappointment. It is that clients make expensive, multi-year decisions on the strength of tools that do not yet exist at the scale promised, and then find themselves locked into a relationship that under-delivers on the very thing that won the business. When that happens, trust erodes, and the next review becomes even more adversarial.
There is a reputational cost too. The advertising industry has spent years trying to rebuild client confidence after earlier controversies over transparency and media rebates. A fresh wave of AI overpromising threatens to reopen that wound, casting the pitch process as a game of who can bluff most convincingly rather than who can actually do the work.
What it means for Australia
The Australian angle here is not incidental. This is a comparatively small and tightly connected market, where the same clients, agencies and consultants circulate through review after review. Reputations travel fast, and an agency that wins on an AI promise it cannot keep will find that word gets around. That intimacy cuts both ways: it makes bluffing riskier, but it also means the culture of overpromising can spread quickly once it takes hold.
The timing matters as well. Australian marketers are under real pressure to show productivity gains from AI, echoing the broader national conversation about whether the country is genuinely building AI capability or simply buying the hype. Boards and chief executives want evidence that their marketing spend is smarter and leaner, and agencies know it. That demand creates fertile ground for inflated claims, precisely the environment in which a consultant’s scepticism becomes most valuable.
There is also a consumer-facing dimension. Australians are already wary of AI in advertising, with research pointing to distrust of AI-generated ads and unease about how the technology is used. If agencies are overselling AI internally to win business, the same loose relationship with the truth can eventually surface in the work itself, whether through AI-made content that misfires or campaigns built on capabilities that were never quite ready. The credibility problem does not stay behind the pitch-room door.
What happens next
The likely correction is a more demanding pitch process. Consultants are signalling that they intend to interrogate AI claims harder, asking agencies to show working tools rather than concept slides, to name the clients where a capability is already live, and to commit to measurable outcomes rather than vague transformation. Some marketers may start writing AI performance guarantees into contracts, turning a pitch-room boast into a contractual obligation with consequences if it is not met.
For agencies, the lesson is that the AI story is becoming a liability as much as an asset. Underpromising and overdelivering has never been fashionable in a competitive pitch, but in a market where everyone is making the same claims, restraint may soon be the thing that stands out. The agencies that win the next round of business could well be the ones prepared to say what their AI cannot do, and to prove what it can. Whether an industry built on selling the dream can learn to sell the reality is the open question, and Australian marketers will be the ones deciding the answer.
Sources: B&T.
