Your Client Has AI. That Doesn't Mean They Have Better Strategy.
A client arrives at the first workshop with an audience segmentation, a competitor review and three possible positioning routes. The documents are thorough, neatly structured and full of plausible conclusions.
Perhaps they were produced with AI. Perhaps the client's team spent six weeks developing them. It doesn't matter as much as people think.
The agency still has to decide whether anyone is asking the right question.
More material is not the same as better judgement
AI can help a marketing team gather information, explore options and prepare arguments. It can also make the team's existing assumptions more productive.
A client who believes strategy is mainly about data may produce more analysis, more segmentation and more reasons to optimise the familiar. Another who prizes inspiring ideas may generate increasingly attractive territories without improving the evidence behind them. The agency still needs to examine the evidence behind a recommendation.
Neither approach is inherently wrong. Data can uncover the unexpected. Creative intuition can reveal possibilities a spreadsheet will miss. The problem begins when one way of looking at the world becomes the only one the organisation considers legitimate.
AI is very good at producing material inside a frame. It is not guaranteed to tell the user that the frame is wrong.
What changes when clients build AI capability
The GWA KI-Whitepaper 2026 (page 41) found that 64.7% of the 170 surveyed company decision-makers reported a specific person responsible for AI in marketing. The same study found substantial interest in bringing more marketing work in-house.
These are self-reported results from a non-representative survey. They indicate organisational attention to AI; they do not measure strategic skill. The implications for what clients will still buy externally depend on the work and the client.
An internal team can become extremely good at executing work while remaining close to the assumptions, incentives and political constraints of its own organisation. Sometimes that proximity is an advantage. Sometimes it makes certain questions almost impossible to ask.
The value of an outside pattern
Experienced strategists carry patterns between assignments. They remember the client whose category looked unique until it started behaving like another market. They recognise a familiar organisational argument disguised as a consumer insight. They notice when a supposedly obvious audience truth is really an internal belief repeated often enough to sound like evidence.
The advantage comes from recognising a pattern and testing whether it applies somewhere new. The evidence and reasoning still need to hold up.
External distance matters too. Someone inside a company may know that questioning a sacred brand assumption will be politically costly. An agency strategist can consider the possibility before worrying about how to sell it in.
That independence is not automatic. Agencies can be just as eager to please, and an external perspective can be superficial. It has to be earned through the quality of the thinking.
Don't improve the answer to the wrong question
Imagine the client brings three positioning territories. A competent agency can compare them, identify weaknesses and help with making a positioning choice. A better intervention may be to ask why those are the three territories under consideration.
What assumption excluded the fourth? What behaviour are we actually trying to change? Which competitor are we treating as the benchmark, and why? Are we solving a brand problem or an internal alignment problem? What would have to be true for the client's preferred answer to be wrong?
Those questions can be uncomfortable. They can also save months of work built on an assumption nobody tested.
The agency may need to offer a different definition of the problem before anyone improves the presentation.
Transparency matters, but it isn't the whole relationship
On page 46 of the same GWA study, 92.9% of the 170 company respondents said transparency about agency AI use was very or rather important. Agencies should be able to explain where AI contributed, how information was handled and who owns the judgement behind the recommendation.
That doesn't mean a client needs a transcript of every prompt. It means the agency should be accountable for the work, regardless of the tools used to develop it.
If a client asks why a positioning direction is right, 'the model suggested it' is not an answer. Neither is 'our strategist has a good feeling about it'. The agency needs to explain the evidence, assumptions, trade-offs and creative opportunity.
The more capable client raises the bar
As clients become more fluent in AI, superficial agency work may become easier to spot. That is uncomfortable for agencies selling polished research summaries as strategic insight.
For good strategists, it can be an opportunity. Less time explaining basic information. More time discussing the difficult decisions. Less admiration for a deck. More scrutiny of what the deck actually recommends.
The agency needs to bring a perspective the client has not yet considered and the judgement to decide whether that perspective deserves to change the work. That judgement has to earn its place in the recommendation.
For strategists working with Mirel, the responsibility remains to test the evidence, challenge the framing and own the final recommendation.