“We can’t afford to turn work away.”
It is the most expensive sentence in professional services, and it is almost always said by someone doing careful arithmetic. The pipeline is thin. The bench is paid whether or not it is busy. A mandate in hand is revenue; a mandate declined is nothing. On those terms the sentence is obviously correct, and firms act on it every week.
The arithmetic is correct and the conclusion is wrong, because the sentence contains a hidden assumption: that the deals available to you next quarter are independent of the deals you accept this quarter. They are not. Acceptance is not a filter you apply to a fixed population of opportunities. It is an input to what that population becomes.
A firm that accepts everything does not end up with more deals than a firm that declines. It ends up with a different, and worse, distribution of them — and it gets there by a mechanism that has been well understood in economics for half a century.
The loss is never the fee you did not earn
Start with the visible number, because it is the one that misleads.
When a firm declines a mandate, the loss it feels is the fee it will not bill. That number is easy to name, it is large, and it arrives at the exact moment the decision is being made. Everything on the other side of the ledger is diffuse, deferred, and invisible in any management account.
The first invisible cost is the bench. A mandate does not consume a budget; it consumes people, and the people it consumes are the senior ones. Worse, the consumption runs backwards to intuition. A clean engagement moves through its phases and asks for partner time at the points where partner time is designed to be spent. A mandate that cannot close asks for partner time continuously, because unresolvable problems do not resolve — they generate meetings. The engagement that will never reach a term sheet is not the one quietly costing you a little. It is the one filling your senior calendar.
The second invisible cost is displacement. Bench capacity spent is bench capacity unavailable, and the mandate you cannot staff next month because this one is still consuming the team is not recorded anywhere as a loss. It simply never appears.
Neither cost shows up in the line the partner is looking at when they say the sentence. Only the fee does. So firms optimise the one number they can see, which is the same error the sponsor makes in a different domain: choosing the visible variable and letting the invisible one decide the outcome.
You are not receiving a random sample
The third cost is the one that changes the argument from bookkeeping to mechanism, and it is where the sentence actually fails.
George Akerlof’s 1970 paper on quality uncertainty described what happens in a market where one side knows the quality of what is being traded and the other does not. Buyers, unable to distinguish good from bad, price at the average. Sellers of good stock find the average price unattractive and withdraw. Average quality falls, the price falls with it, and more good sellers exit — a loop that runs, in Akerlof’s account, until “the bad drives out the good.”
Advisory mandates are traded under exactly that asymmetry. The sponsor knows things about the deal that you will not learn for six weeks: which lender has already passed, what the last adviser said before resigning, whether the licence is what the letter says it is. You are pricing on averages. They are not.
Now add the part that makes it bite. Sponsors do not approach advisers at random, and they do not accept terms at random. A sponsor with a clean, well-evidenced project has options, negotiates hard on scope and fee, and takes time. A sponsor whose project has already been refused twice has fewer options, negotiates less, and accepts faster.
Readiness to accept your terms is negatively correlated with the quality of the underlying deal. The mandates that arrive easiest are, on average, the ones the market has already looked at and declined.
Petroleum engineers know the same shape from the licensing round: Capen, Clapp and Campbell’s 1971 finding that in competitive bidding under uncertainty, the winner is systematically the bidder who most overestimated the prize. Win the auction and you have learned something unwelcome about your own estimate. Win a mandate on the sponsor’s first ask, without friction, and you have learned something similar.
None of this is a claim about sponsors’ honesty. It is a claim about selection. Good deals are refused less often, so they reach you less often per unit of eagerness.
Declining is a signal, and signals compound
Here is the loop that the opening sentence ignores.
What a firm accepts becomes known. Not through announcements — nobody publishes their declines — but through the ordinary traffic of a market: which deals a firm was seen working on, what the counterparties experienced, what the lawyers and bankers and DFI officers say when a name comes up. Before long, a market has a working view of what any given adviser will take on.
That view then determines what is brought to them. A firm known to apply standards receives deals that have been pre-filtered by the referrer, because no lawyer wants to send a client into a refusal. A firm known to take anything receives everything, including everything that has already failed elsewhere. The reputation does not merely reflect the intake; it produces the next one.
Call this the Selection Loop: every acceptance decision is a marketing decision that pays out later. It is the reason decline discipline is not a cost centre wearing a virtue costume. It is the cheapest pipeline-quality instrument a firm has, and the only one that operates while nobody is working on it.
The corollary is uncomfortable for firms in a thin quarter. The mandates you accept when you are desperate are the ones that will define what you are offered when you are not.
Decline on named grounds, not on feel
A decline discipline that lives in a partner’s instincts is not a discipline. It is inconsistent between partners, unexplainable to the referrer, and impossible to defend when a junior asks why this one and not that one. It also collapses precisely when it matters, which is when the quarter is thin.
The remedy is a written set of grounds, and the professional standards world reached this conclusion long before the advisory world did.
| Ground | What it protects | The tell |
|---|---|---|
| Counterparty and source-of-funds failure | The firm’s licence to operate | Unwillingness to answer ownership questions plainly |
| Sector or product mismatch | Delivery quality | The work requires a capability you would have to invent |
| Capability gap | The client | You are the second-best team available and know it |
| Conflict of interest | Both clients | The obvious counterparty is already yours |
| Insufficient engagement economics | The bench | The scope only works if nothing goes wrong |
| Reputational or political exposure | Every other mandate | You would hesitate to name the client to a lender |
| Rejection of the engagement charter | The engagement itself | The sponsor wants the outcome without the process |
International Standard on Quality Management 1 requires firms performing audits and assurance work to establish quality objectives so that judgements on whether to accept or continue a client relationship are appropriate, based on information about “the nature and circumstances of the engagement and the integrity and ethical values of the client… sufficient to support such judgments”, and on “the firm’s ability to perform the engagement in accordance with professional standards” (¶30(a)).
Then it adds the paragraph that matters most here. The standard requires that “the financial and operational priorities of the firm do not lead to inappropriate judgments about whether to accept or continue a client relationship or specific engagement” (¶30(b)).
Read that again in the light of the opening sentence. An international standard-setter, writing for firms of every size, considered it necessary to state explicitly that needing the revenue is not a reason to take the work. They did not write that because the problem is rare.
“We are too small to say no”
This is the honest objection, and it is not answered by telling a small firm to be braver.
Take it seriously and it inverts. The smaller the bench, the larger the share of it that a single bad mandate consumes, and the longer the recovery. A firm of forty can carry a mandate that will never close; a firm of six cannot, because that engagement is a third of its senior capacity for two quarters and there is no second team to keep the good work moving. Decline discipline is not a luxury that arrives with scale. It is most valuable precisely where the capacity is thinnest — which is why ¶30(b) is addressed to every firm, not to the large ones.
There is also a false binary hiding in the objection. The alternative to accepting is not refusing; it is routing. A business too early for institutional structuring is a real business with a real need, and it can go to an affiliate, a specialist, or a named contact with a warm introduction. The firm keeps the relationship, the sponsor gets served by someone who can actually help, and the referrer sees a firm that solves problems rather than one that simply says no. Very few declines need to be refusals.
And a decline made on a named ground, delivered quickly and without a lecture, is a better commercial act than a slow yes. Sponsors remember the adviser who told them in a week that this was not fundable yet and what would change that. They rarely come back to the one who took the retainer and told them the same thing in month five.
Back to the sentence
“We can’t afford to turn work away” treats next quarter’s pipeline as a given and this quarter’s revenue as the only variable. It has the causation backwards. The pipeline is the dependent variable, and what you accept is what sets it.
A firm’s decline rate is not a measure of the business it is losing. It is the mechanism selecting the business it will be offered.
So the first move is not a policy document. Take the last five mandates you accepted, and beside each one write which of the grounds above you could have declined it on, and whether anyone raised it at the time. Then take the last three you declined, and write what happened to your pipeline in the two quarters after. Most firms have never looked at those two lists together, and they are the only evidence that settles the argument.