Coca-Cola's New Agency Brief Skips the Step That Genentech's CMO Does First
CMOs know exactly what they want from their next agency. Most of the time, that clarity is the problem.
Coca-Cola CMO Manolo Arroyo went on record this week saying the brand's next agency appointment needs two things: creator media chops and retail media capabilities. That's a confident, specific brief. The confidence is the tell.
Not because the requirements are wrong. Creator and retail media are both genuinely important surfaces in 2026. The question is what diagnostic process produced those requirements. Because there is a version of this story where Arroyo ran a perception audit, identified that Coca-Cola's brand messaging is losing ground specifically on creator-native platforms and at the retail point-of-decision, and concluded that the capability gap is real and addressable by agency selection. And there is a version where those are simply the hottest capabilities in the market right now, which means every CMO is including them in every brief, which means the brief tells you more about the zeitgeist than about Coca-Cola.
From the outside, these look identical. From the inside, only one of them leads to a better outcome.
The Capability Brief Versus the Diagnostic Brief
A capability brief answers: what can this agency do? A diagnostic brief answers: what is our actual problem, and what capability would address it?
Most agency searches in practice run the capability brief first and treat the diagnostic as something the agency will help figure out. That sequence is backwards. When you hire for capability before completing the diagnosis, you are selecting for an agency's ability to execute a solution you haven't fully defined yet. The agency, rationally, will apply the capabilities you hired them for, whether or not those capabilities fit the actual problem.
This is not a knock on Arroyo or on Coca-Cola. It's a structural observation about how capability briefs work. The brands that consistently get more from agency relationships tend to be the ones that arrive with a sharper problem statement, not a sharper list of desired capabilities. The capability requirements should be output from the diagnostic, not input to it.
Genentech CMO Zoë Lazarre offered something closer to the diagnostic model in a recent McKinsey conversation. Her frame was not about which channels or capabilities the marketing organization needed. It was about why growth depends as much on culture as technology, and why patient trust, once you understand what actually erodes it, shapes every other decision downstream. She is working backwards from a clearly articulated problem. The capability decisions follow.
What Behavioral Science Says About Checking the Problem First
A piece published this week at BehavioralEconomics.com makes the diagnostic point cleanly with a policing example. When residents report feeling unsafe in a neighborhood, the instinct is to address the perception, launch a reassurance campaign, increase visible police presence, issue communications about crime statistics. The behavioral insight is that this entire response can be a mistake if you haven't first determined whether the perception reflects a real problem or a misread one.
If crime is actually up, the perception is accurate and the fix is the crime rate, not the communication. If crime is stable but visible disorder has increased, fixing the disorder fixes the perception without any messaging. If the perception is concentrated among a demographic that has had direct negative interactions with police, communications about aggregate crime statistics won't move it at all. Three different root causes, three different interventions, and the surface presentation (residents feel unsafe) looks identical across all three.
The brand strategy parallel is direct. If consumers perceive a brand as disconnected or low-value, the instinct is to address the perception. Hire for creator media to reach consumers where they are. Build retail media presence to be visible at the point of purchase. Issue more content. The question that almost always gets skipped: does this perception reflect a real product or experience gap, or is it genuinely a messaging and distribution problem?
If the answer is a real product gap, then improving distribution of the current message amplifies the wrong signal. If the answer is a trust gap driven by how consumers experience the brand at the service layer, then creator media won't touch it. If the answer is a relevance gap among specific age cohorts who have simply moved on, then retail media presence at the decision point still doesn't close the underlying relevance deficit.
Running the diagnostic before issuing the brief means arriving at a much more specific question. Not "what agency has creator and retail media capabilities?" but "what agency understands the specific perception-to-reality gap we're addressing and has capabilities relevant to that gap specifically?"
The Genentech Approach: Starting With Patient Trust
Lazarre's framing deserves more attention than it usually gets in CMO coverage, because she is doing something structurally different from most of the brand marketing conversation this year.
Her emphasis on culture as a prerequisite for technology suggests she has diagnosed the root cause of brand trust erosion in pharma as an internal problem before it becomes an external one. When a pharma brand loses patient trust, the mechanism is often not messaging failure. It is a gap between what the brand communicates about its commitment to patients and what patients actually experience in the interaction with the product, the healthcare provider, the billing process. Closing that gap with better messaging does not close the gap. It often makes it worse because it amplifies the contrast.
So the diagnostic question for a pharma CMO is: where does the experience break? And the answer to that question determines whether marketing is the right intervention at all, or whether marketing is downstream of a product, access, or service problem that needs to be addressed first.
This is why brand trust is operational before it is aspirational. The channels and capabilities you bring to the marketing function should be answers to that operational question, not the question itself.
Lazarre's framing on AI is similarly instructive. She is not describing AI as a capability she needs to build for its own sake. She is describing it as a tool for understanding what patients actually need at different points in their journey, with the implication that understanding produces more relevant action. The tool serves the diagnostic, rather than the diagnostic serving the tool.
The Portfolio Construction Parallel
There is an interesting parallel in the personal finance data published this week by Of Dollars and Data, which analyzed the best-performing portfolio over the last 50 years. The finding is that diversification into asset classes that were performing well in any given decade rarely produced superior long-run outcomes. The investors who fared best over the full period tended to hold consistent thesis-driven allocations rather than optimizing for whichever surface was hot at any given moment.
The CMO capability brief has the same structural problem as trend-following portfolio construction. Creator media is performing well right now. Retail media is growing. Including both in your agency brief feels like sound diversification. But this is the logic of buying what's recently worked, not of building toward a thesis about where your specific brand problem lives and what capabilities address it.
The original analytical point here is this: in both portfolio management and brand strategy, the most expensive bias is allocating to distribution of a current position rather than diagnosing whether the position is correct. A portfolio that doubles down on last decade's winners and a CMO that doubles down on currently-hot channels are making the same category error. They're both optimizing execution before validating the thesis.
The brands that pull ahead over a five-year horizon are not usually the ones with the best creator media agency. They're the ones that correctly diagnosed what their consumer relationship problem actually was, and then hired for that specific problem.
The Diagnosis Sequence CMOs Should Run First
Before writing a capability brief for any major agency search, the diagnostic sequence runs roughly as follows.
Map the actual perception gap
Start with what consumers believe about the brand, then test whether each belief reflects a real operational or product truth. The confidence trap in brand strategy is assuming that a perception is purely a messaging problem, when it often reflects an experience that the messaging is contradicting. A consumer who believes your brand doesn't understand them is usually pointing to something real.
Segment the perception by source
Consumer perceptions are almost never homogeneous. The residents who feel unsafe in the policing example aren't all feeling unsafe for the same reason, and the fix for each subgroup is different. Similarly, a brand perception problem among Gen Z consumers on creator platforms has a different root cause than the same nominal perception among older consumers encountering the brand at retail. The capability requirements for those two problems are not the same, even if the surface measurement looks identical.
Test whether marketing is the right intervention layer
This is the question most CMOs are institutionally prevented from asking, because the answer might be "the product needs to change" or "the service experience needs to change" and that is outside the CMO's authority. But issuing a capability brief before asking this question means you might be hiring an agency to execute very well on a problem that marketing cannot actually solve. The CEO trust literature makes this same mistake at a higher level, assigning trust-building responsibility to communications functions when the trust gap is operational.
Build the capability brief from the diagnosis output
Only after those three steps does it make sense to write a capability brief. At that point, the brief becomes much more specific. Not "we need creator and retail media capabilities" but "we have a diagnosed relevance gap with 18 to 24 year olds that is driven by low presence on short-form video and is a genuine distribution problem, not a product problem, and we need an agency that has built and measured creator programs in that cohort specifically."
That brief is harder to write. It requires doing the diagnostic work first. But it produces a much better agency selection, and it means the agency you hire is working on the right problem from day one.
Branding Strategy Insider published a separate piece this week arguing that the broader consumer moment is defined by a loss of agency. Consumers feel like things are out of control and they want more influence over their relationships with brands. That finding lands differently depending on where the diagnosis starts. If you read it as a distribution signal (consumers want to find brands in more places, control the channels they use), you build creator and retail media capability. If you read it as a trust signal (consumers want to feel like the brand actually responds to what they tell it), the capability you need looks completely different.
The perception audit is what tells you which reading is correct for your specific brand and your specific consumer.
If you're working through a brand strategy or consumer perception problem and want a research-grounded take, the STI research tools are built for exactly this kind of diagnostic work.