I was sitting in a weekly operations review with a grain trading client (yes, I know — SF strategic finance guy + grain trading is a tale as old as time). The trucking side of the business had quietly gone underwater for the week. When I asked why, the logistics manager explained that they’d run some outside loads for one of their transloading customers. This customer routinely takes a full trucking day of capacity due to delays, but we get paid on volume instead of hours, which is what put the loads underwater. The team’s solution, already half-formed, was to bring on a third-party hourly trucking vendor to absorb the volatility.
I asked a different question: how much would it hurt the P&L to drop them?
The customer generates around $16,000 a year in free cash flow, enough to be impactful, particularly for a low-margin business like commodities. Once you walk through the rest of what comes with them, the picture quickly shifts. Their railcars arrive with ours, competing for yard space, which produces demurrage when prioritization gets confused. The full-day trucking commitment displaces revenue from the grain side, the side that actually produces economies of scale. We realized the operational burden of running outside loads for transloading customers, with its variable volume and non-standard specifications, is a fundamentally different business than the one we’re trying to be the best in the world at.
The real question on the table was never about the $16k. It was: are we a grain trading business, or a transloading business? The answer needs to govern every staffing decision, capital purchase, onboarding standard, and operational priority. The team was building two businesses without picking which one was the real one. The underwater trucking week was the symptom.
We dropped the customer. The conversation happened recently enough that the contract is still winding down — but the math was unambiguous, and the underwater trucking week made the case the team needed. The $16,000 of free cash flow was the cheapest line item the business was carrying.
A calibration note before the framework. I’m writing for the kind of $2–20MM founder whose default is yes: too many initiatives, customers, and product lines; filling their calendar full of work that doesn’t scale their business. If your default runs the other way — if you’re sitting on capital you should be deploying, killing experiments before they have a chance to find product-market fit — the same framework, applied honestly, would tell you to say yes more often. The True Cost of Yes math is calibration-dependent. For most of the founders I work with, the calibration error runs in one direction. If you’re the exception, this article is the wrong medicine.
For the chronic-yes founder, saying no is capital allocation. Every yes consumes a finite stock of operator time, team attention, working capital, and strategic clarity. Treating it as the allocation decision it actually is, rather than as a customer service question, derives more value from your time and builds your business faster.
The True Cost of Yes
The framework I use with partners is simple to state and unforgiving in practice:
True Cost of Yes = Direct Cost + Opportunity Cost + Complexity Tax
The first two are familiar enough that most operators eventually find them. The third is where the damage gets done.
Direct Cost
The line items your accountant already tracks. Variable costs of fulfilling the yes: labor, fuel, materials, hosting, software seats, shipping. In the trucking example, the cost of the driver and fuel for the outside loads.
Most P&Ls show this clearly and most founders calculate it correctly. It’s also the smallest of the three costs for any yes that lives inside an existing business. If the only cost of a deal were variable cost, every deal that priced above variable cost would be worth taking. That math is wrong, which tells you the math is incomplete.
Opportunity Cost
What your scarce resources could have been doing instead. The trucking day given to that outside customer was a day not spent moving grain. The hour the founder spent reviewing the unusual contract terms was an hour not spent on the core sales pipeline. The shop floor capacity allocated to a custom run for one customer was capacity not allocated to your highest-margin standard SKUs.
To price opportunity cost honestly you need two things: a clear sense of what your scarce resource is, and a current price for it. Most $2–20MM businesses have at least three scarce resources, ranked roughly: founder time, key technical or operational staff capacity, and working capital. Trucking businesses add a fourth: rolling stock and yard space. SaaS businesses add a fourth: engineering hours. Lifestyle brands add a fourth: inventory dollars in your best SKUs.
The price for each is the gross margin your scarce resource produces when allocated to your highest-return activity. In the trucking case, an hour of capacity allocated to grain produces gross margin X. An hour allocated to the outside transloading customer produces something lower. The difference is the real opportunity cost of every hour spent on the customer who looked like found money. Almost every time I've run this, the opportunity cost alone is bigger than the FCF. That's before the complexity tax.
Complexity Tax
The hidden one. The cost that doesn’t show up on a P&L line until well after the decision has been made. It has four components — operations are impacted by the first three and ultimately this creates the most expensive tax, strategic ambiguity.
Operational mode mismatch. Different customers and product lines run on different cadences, specifications, and tolerances. Mixing variable-volume work alongside fixed-schedule work isn’t just operationally annoying. It forces your team to maintain two mental models for the same task, which doubles the error rate at the seams.
System contamination. One non-standard customer means one non-standard process, which means one rule for them and a different rule for everyone else. Multiply across the team and your standards stop being standards.
Management overhead. Every yes generates recurring decisions: which crew, which equipment, which pricing tier, which escalation path. The founder who said yes is the one who keeps getting pulled in to adjudicate the exceptions. Multiply across a portfolio of marginal yeses and the founder becomes a switchboard.
Strategic ambiguity (what the first three accumulate into). Every yes that lives in a different operating mode quietly raises the question of which business you’re actually running. The grain operation I described had been building two businesses. Three or four of those yeses, accumulated over a few years, and the answer to “what do we do best in the world” gets fuzzy enough that the team starts hedging both. Fuzzy strategy compounds backward.
Run the formula on the trucking customer. Direct cost ate most of the variable revenue, making the contract near break-even before anything else. Opportunity cost on displaced grain margin — based on grain trading unit economics, a trucking day allocated to grain is worth meaningfully more than one allocated to outside loads — added multiples of the FCF in foregone earnings. Operational mode mismatch and management overhead ate hours of the logistics manager’s week. System contamination showed up directly as demurrage on jammed yard space and as delayed grain revenue from prioritization confusion. And the largest piece, the strategic ambiguity, was seen in the fact that the team’s instinct was to add more complexity via a third-party trucking vendor rather than question whether the customer should exist.
Add the three currencies honestly and the $16,000 of free cash flow was the most expensive $16,000 the business had on its books.
Complexity Is Capex of Time
This is the framing I find myself repeating in client meetings until I’m tired of hearing it.
When a founder buys a piece of equipment, they understand it as capital expenditure. They calculate the payback period. They compare against alternatives. They make a deliberate decision because cash is visible and finite.
When that same founder takes on a new customer, a new product line, a new channel partnership, or a new internal initiative, they almost never put it through the same analysis. The cost shows up as time and attention rather than dollars, and time and attention feel more elastic than they actually are. They aren’t. The operator-hours available in a year are a fixed stock. Allocating them is the same decision as allocating dollars, with the same compounding consequences.
Where the value of no actually lives: the highest-return use of operator time in most $2–20MM businesses is compounding the existing core. Optimizing the sales engine you already have. Improving the margin on the SKUs you already sell. Deepening the relationship with the customers who already love you. These are the bets you can size confidently because you already know the unit economics. New bets, by definition, have unknown unit economics, which means you’re paying a learning premium on top of the direct cost.
The right time to say yes to a new initiative is when the existing business is so well-optimized that the marginal hour of operator time produces less return there than it would on the new thing. That bar is rarely met, and almost never met as early as founders convince themselves it is.
Where It Shows Up
The two places where saying no produces the most reliable returns are the wrong customer and the wrong distribution strategy.
Every customer base has a bottom decile that costs more than they pay. Some of the cost is direct. Most of it is complexity tax. The non-standard customer who demands custom terms, custom delivery, custom packaging, or custom support eats into a team’s ability to standardize, which is the only way a small business builds real operating scale. For most clients, the bottom decile nets flat or slightly negative once fully loaded, and the capacity it consumes would produce 2–3x its revenue if reallocated to the top quartile. It may feel like shrinking to drop 10% of your most operationally painful revenue, but that capacity will let you compound the rest of the business more effectively by reallocating it toward expanding or adding simpler customers. What looks like shrinking is reallocation.
I run the same playbook I preach, in a softer form than the trucking client could use. They said no unilaterally. I price work at its true cost and let the buyer decide. Both refuse engagements that don’t compound the core. Theirs is the stronger version. Mine is the weaker version, available to a service business where pricing is the lever I have.
An investment model build for an early-stage company raising venture capital can be a real piece of revenue, but it produces no operational repeatability for the business I’m trying to build, which is a finance partnership that compounds value for owner-operators over multi-year engagements. So I price those engagements high enough that the proposal absorbs the opportunity cost and the complexity tax. A recent $25,000 model build for a very large venture capital raise was rejected at that price. If they were my ICP — an owner-operator with a long view on their business — I could have priced in the long-term value of the relationship. Since it was a one-shot venture raise, it wasn’t worth trying to win the deal. The rejection let me spend the next six weeks on productizing the parts of my practice that actually scale: standard analytics templates, repeatable diligence frameworks, the operational infrastructure for adding the next finance partner. None of that work happens if I take the engagement at half the number, even if I’m making my target rate. The reps I would have gained were the wrong reps for the business I’m building.
The distribution version of the same problem is overseas expansion. The existing brand and product travel, the thinking goes, so why wouldn’t we serve a larger TAM. The unspoken cost is two to four weeks of executive time spent finding distribution partners, meeting regulatory requirements, working through trade barriers, and managing currency exposure. That bandwidth, allocated to the domestic business, would have produced a measurable lift in a known operation. Allocated to international, it produces an uncertain bet on a new operation. The expected return on the domestic bet is almost always higher, especially in the first $5–15MM of revenue where the core is still under-optimized.
Three Patterns I See Constantly
Three patterns I see constantly. Each is a different shape of yes whose true cost — opportunity, complexity, or both — arrives after the decision is made.
“We Should Be Doing This”
Most founders I work with keep a running list of things they “should be doing.” A new channel. A new market, feature, or partnership. The list grows faster than the team can execute against it. Each item, considered individually, sounds reasonable. The list considered as a whole is a slow-motion strategic identity crisis. What the founder is actually saying, when they bring up the list, is that the existing business doesn’t feel like enough. Sometimes that’s a real signal. Most of the time it’s a story problem rather than a business problem, and the answer isn’t to add another initiative to the list. The answer is to articulate what the existing business is, why it’s enough, and why this particular adjacency is the wrong fight to pick this year.
Commission Complexity
The pattern I see in $5–15MM businesses with a real sales team is the founder trying to engineer the perfect commission structure. They want it to track gross margin, factor in product mix, adjust for customer tier, weight for new vs. renewal, and account for territory. The math is elegant. The result is a sales team that spends an hour a week on a personal commission calculator instead of selling. Salespeople are motivated by simple, legible incentives. A commission plan a rep can recompute in their head produces more revenue than one that requires a spreadsheet. The yes here is to internal complexity that feels like good operating practice and functions, in practice, as a tax on the sales engine.
Sunk Cost Survival
The hardest no, by a wide margin, is the one for a product line or customer relationship the business has invested in for years. The founder remembers the early version of it, the team that built it, the customers who came in through it. The fact that it now consumes more in management attention than it produces in contribution margin is hard to see clearly, because the line still throws off some revenue and still has some customers and still feels like part of who the company is. The reframe I use: every quarter you choose not to wind it down is a fresh decision to allocate this quarter’s operator time to it. The test isn’t whether the line was a good decision originally. It’s whether you would start it today knowing what you now know. If the answer is no, the next question is when you stop.
Before your next strategy meeting, spend a week subtracting before spending a quarter adding. Walk the customer list, the product line, the channel list, the open initiative list. Run the True Cost of Yes math on each. Drop what doesn’t compound the core. Pre-allocate the freed capacity to a specific compounding use before anything else can refill it.
Businesses that do this end up smaller in line count, larger in revenue, materially better in margin, and far easier to operate. The founder’s calendar opens up. The strategic question stops being “what do we add?” and starts being “what do we make better?”
That’s a different kind of growth than the one most founders are sold. It is growth. It just doesn’t put you on a panel.
If you’re somewhere in the middle of a yes you suspect is the wrong yes, I’d be happy to help you think through it: duncan@saorsapartners.com
As requested by the audience, this piece was shorter by design. If you prefer longer articles, or more theoretical pieces, please let me know in the comments below. If this is your first time reading, please subscribe to Conduit of Value to get it in your inbox every week.