This article is based on written responses from Underscore CEO Lauren Boyer to questions from Greg Kihlström for The Agile Brand Guide’s Expert Mode series.
The agency question looked settled at the end of 2025. Omnicom closed on Interpublic, the combined company cleared $25 billion in revenue, more than 4,000 jobs went, and names that had been on doors for decades — DDB, FCB, MullenLowe — stopped existing. Scale won. For a CMO planning a roster review, the practical read was that consolidation makes the buying side simpler: fewer contracts, better rates, one number to call.
Lauren Boyer runs a company that would seem to have every reason to argue the opposite. She’s CEO of Underscore, a media agency built for biopharma launches in rare disease and oncology, and specialization is the thing she sells. She knows it, too — she says so on the record. What’s interesting is that her argument isn’t really about big versus small. It’s about a cost most marketing organizations are already paying and almost none of them are counting.
You’re Paying for the Learning Curve Either Way
The depth-versus-leverage frame is the one procurement uses, and it invites an agency leader to claim that outsiders can’t learn the category. Boyer won’t make that claim. She thinks capable people can learn almost anything, which means the defense of specialization has to rest somewhere else. So she moves the question from what the agency knows to who funds the gap.
I don’t think the case for specialization is that a strong generalist cannot learn a category. Smart people can learn almost anything. The real question is whether the client should be paying — in time, risk, and executive attention — for the agency to learn what should already be built into its judgment to maximize brand impact.
The bill she’s describing lands on the client’s calendar, not the client’s invoice, which is exactly why fee comparisons miss it. She names three forms it takes: meetings spent teaching fundamentals, recommendations that sound right but miss the operating reality, and early performance signals that get overread or ignored. None of those appear in a cost-per-hour comparison across a shortlist. All of them are opportunity cost borne by the most expensive people in the building. There’s a practical move in this for any marketing leader running a competitive review: ask each incumbent to write down what it had to learn about your business in its first six months, and who did the teaching. The agencies that can answer honestly will tell you something. The ones that say there was no learning curve are telling you something too.
Automate the Task, Then Rebuild the Apprenticeship
The standard AI story in agency land is that the technology absorbs the work junior people used to do, and everyone books the margin. Boyer’s problem with that story isn’t the automation. It’s the second-order effect nobody’s budgeting for.
This is one of the most important questions agencies should be asking about AI, because the easiest version of the story is also the most dangerous one: use AI to absorb junior work, celebrate the efficiency, and then realize five years later that we stopped forming senior thinkers.
She isn’t arguing for keeping manual work around out of sentiment. If a task doesn’t require human judgment, automate it — but then rebuild the apprenticeship on purpose, because the old one worked by proximity. Junior people learned by compiling, checking, formatting, and sitting near decisions long enough to absorb how they got made. Strip that out and the learning and development path has to be redesigned to put new people next to judgment much earlier: what question are we actually answering, what does the data not know, what would change the recommendation. Her guidance to clients on fees is the useful part here, and it cuts against the obvious assumption. Don’t expect automatic discounts. Expect to be able to point at something that’s faster, sharper, or more useful than it was last year. If your agency says it’s using AI and you can’t name that thing, press harder. It’s a fair question to put in a quarterly business review, and a pretty revealing one.
A Specialization Story That Doesn’t Run in a Straight Line
Underscore was pitched to us as an agency that stayed specialized while the industry consolidated around it. The record is bumpier than that. Boyer has previously described the agency focusing on health and wellness from 2009 to 2017, then broadening to serve any vertical, and today Underscore presents itself as built exclusively for complex biopharma. Asked about the gap between the story and the history, she didn’t smooth it over.
I would not describe our path as perfectly linear, and I would rather be honest about that than try to retrofit a cleaner story. We’ve been around for nearly 25 years and started more broadly. We have continued to narrow our focus, and I am glad we did, because it taught us the difference between transferable capability and true differentiation.
That distinction does real work for anyone buying agency services. Capability travels between categories; differentiation mostly doesn’t, and an agency that can serve your category isn’t the same as one whose team asks better questions in it. Boyer also told us that rare disease has touched her own family, which she says shapes how she thinks about small patient populations deserving the same investment as large ones. Worth saying plainly: she has a commercial stake in every word of this, and she volunteered the boundary anyway. If a global CMO is consolidating creative, media, commerce, production, analytics and technology under one operating model across many markets, she’d point them at a holding company — a different business problem, in her framing, not a lesser one. And her own disqualifier is sharper than most agency leaders will give you. If complex biopharma commercialization isn’t among the hardest problems on a client’s table, she says that client probably shouldn’t have Underscore in the consideration set at all.
Logos Prove Exposure. Depth Sounds Different.
Most roster reviews open with the wrong question. Boyer would replace “are we satisfied with our agencies” with something closer to “where are we asking each partner to solve problems they weren’t built to solve,” on the grounds that satisfaction is too soft a measure to act on. Then there’s the artifact at the center of every pitch, the one she says fools people most reliably.
The thing that looks like specialization but often is not is a credentials deck full of category logos. Logos prove exposure. They do not prove depth.
Her replacement test is auditory rather than documentary, and it’s checkable in a room. Real depth shows up in the questions an agency asks: how much of the firm actually works in this category, how long the people leading the account have been doing this work, where conventional wisdom in the category breaks down, where the data is thin or delayed or easy to misread. For the senior-involvement promise that every agency makes, she offers two questions a CMO can ask in the pitch itself — who will be in the work with us every week, and can I meet them now; and, after the first 90 days, what decisions will you personally help us make. The second one separates executive sponsorship from actual contribution. She’d also push clients to talk with reference accounts without the pitch team running the call, and to ask those clients whether the senior people from the pitch are still involved a year later. That answer is usually short and usually honest.
The consolidation headlines made this feel like a structural question — pick a side, big or small, and live with it. It isn’t. The decision in front of most marketing leaders is narrower and more boring than that: which of the problems you’re facing in the next two years is each partner genuinely built for, and which ones are you paying someone to learn on your time. Before the next review, write down the three hardest commercial problems coming at you. Then look at the roster and ask which agency was hired for a problem you no longer have.