There is a specific kind of mistake that agency founders are built to make. A deal arrives — large enough to reshape the quarter, complex enough to be interesting, high-profile enough to feel like validation. Every instinct says take it. The revenue is compelling. The work is real. The client is serious. Saying yes is what growing founders do. It is part of the identity.
In late 2019, a deal like that landed on my desk at BBO, and I spent three weeks trying to make it work before the board stopped me. They were right. Understanding why they were right took me longer than it should have.
The terms beneath the headline
A travel-industry company wanted a full-service digital marketing partner — paid search, paid social, SEO, conversion optimization, analytics. They had found us through a referral and liked that we could handle multiple channels under one roof. The brief was ambitious in the way that makes agencies lean forward in their chairs.
The problem was the financial structure. The proposed contract placed a disproportionate share of business risk on us: significant upfront investment in tooling and staffing before revenue recognition, payment terms stretched past what our cash-flow model could absorb without strain, and performance clauses that would leave us carrying costs with no recovery mechanism if triggered. We offered two alternative structures to reduce our exposure to manageable levels. Both were declined.
The counterparty wanted a partner who would absorb business risk in exchange for the upside. Not unreasonable in principle — but the asymmetry was sharp. Their worst case was switching agencies. Ours was six figures of unrecoverable cost.
What the board saw
I went to the board half-convinced we should find a way. The revenue was real, the client was notable, and I had already invested three weeks of attention in getting to this point. That investment was shaping my judgment in ways I did not fully recognize — I was evaluating from a position of sunk cost, not strategic clarity.
The board cut through it in under an hour. They read three things I had been treating as separate concerns as a single picture. Our exposure if the engagement ended at the worst possible moment was uncomfortable — we would need to hire at least two people specifically for this account, investing in capacity before revenue covered the ramp-up. At steady state, this single client would represent over twenty percent of our total revenue — structural concentration risk regardless of how good the relationship was. And redirecting senior capacity to onboard an engagement of this complexity would degrade service to existing clients during the transition.
Any one of those was manageable in isolation. Together, they described a deal where the attractive headline number was attached to a risk profile that did not match our current position. We were not in survival mode. We did not need to bet the balance sheet on a single engagement, and the potential upside did not compensate for the specific combination of risks this structure carried. The board logged it as a governance decision, not an operational one. That distinction mattered — it meant the decision lived in the governance record, not in my inbox. It was institutional, not personal. I could not revisit it in a weak moment at 2 AM without reopening a formal process.
The arithmetic of no
For two days I second-guessed the decision. Then I looked at the pipeline and the math became obvious. The three weeks I had spent evaluating, modeling, and negotiating this single opportunity could have gone toward four or five standard engagements with conventional terms and no structural exposure.
Large deals consume disproportionate organizational energy. Not just time — mental bandwidth, senior attention diverted from delivery, smaller opportunities that do not get pursued while the team is oriented around the shiny object. The opportunity cost of chasing a deal that is too large for your operating model compounds faster than most founders appreciate.
I went back through my calendar for those three weeks and counted the meetings: eleven sessions dedicated to this one prospect, plus the preparation time around each. Eleven slots that could have advanced four or five other conversations to signed contracts. The deal I did not close cost me the deals I did not start.
The second realization was more personal. I noticed, with some distance, how much of my push toward yes had been identity rather than analysis. Landing large clients is what growing founders do. The pull was not coming from the spreadsheet. It was coming from a self-image, and recognizing that was worth more than the deal itself.
The discipline that followed
After that decision, we formalized something we had been doing informally. Before any engagement above a certain revenue threshold, we now run three filters:
Exposure at worst-case termination. Not revenue lost — actual unrecoverable cost. If the number exceeds what we can absorb in a bad quarter, the terms need restructuring before we proceed.
Concentration ceiling. No single client above fifteen percent of total revenue. Hard cap, not guideline. We have turned down scope expansions from existing clients to stay under it.
Capacity dependency. Does the engagement require hires specifically for this account? If so, what happens when the account disappears before those people have been redeployed?
None of this is novel. Most consultancy management literature covers the same ground. The difference is applying it before the negotiation becomes emotional rather than after. That declined deal taught us something fundamental: the moment you begin negotiating terms, you are already psychologically invested in yes. The governance filter has to run first, or it does not run at all. By the time you are arguing about payment terms, the psychological commitment is already made. The filter needs to preempt that.
We have taken on larger engagements since — structured differently. The question was never the size of the opportunity. It was whether the risk was proportional, and whether we were the right party to carry it. The ability to say no is not a luxury. It is a capability, and building it requires a pipeline diverse enough that no single deal feels existential and a governance process strong enough to override the founder’s instinct when the numbers say stop.