Founders ask Collective 54 this 2 times in our records, none of them in 2026. The scope changes, estimating and escalation answers on this site cover handling change requests, estimating hours and raising issues; this page covers the decision once an overrun has happened.
The engagement manager essay in the newer book lists the decisions that only a person can make in an engagement: when to escalate, when to absorb friction, when to push back, when to trade speed for quality, and when to protect margin versus relationship. It says these decisions define the role. As an inference, the way to make this one well is to name the cause of the overrun before deciding who pays for it. Most overruns fall into four groups.
The client changed or added to the work. This is a scope change, and the default is to go back to the client.
The client did not do what the engagement needed from them: late information, slow decisions, missing people. The default is also to go back, with the record of what was agreed.
Your estimate was wrong, or your delivery was slow or needed rework. The default is to absorb it.
The cause is mixed or unclear. The default is an open conversation, often ending in a shared solution.
The essay says most firms lose margin because they treat scope creep as client service, and that great engagement managers treat it as an economic decision. When the client asks for something out of scope, the engagement manager frames the tradeoff clearly, offers options, such as reducing scope, extending the timeline or increasing the budget, and prices and contracts the adjustment for the current engagement. It says to do this without drama, without apology and early, while the client still trusts the firm. The scope changes answer on this site covers the mechanics.
The service design essay asks every service to clarify the role the client must play for the service to succeed. The account executive essay adds that selling is not complete until the buyer is operationally committed and positioned for successful delivery. As an inference, if the client role was written down, a delay on their side is a change to the engagement, and it is fair to discuss its cost. If it was never written down, the conversation is harder, and the fix belongs in the next proposal.
The essay lists the ways margin leaks after a deal closes: scope drift that feels small at the time, rework rationalized as client service, missed handoffs, unclear ownership, overstaffing and slow delivery. As an inference, when the cause is one of these, the client did not create it and should not pay for it, especially on a fixed fee. Absorb it, but do not hide it. Tell the client the work took more than planned and that you are covering it. It protects trust, and a concession the client knows about is worth something; one they never hear about is only a cost.
Then find out why. The essay says most engagement problems are downstream of design problems: services that are underspecified, over-customized or economically unrealistic leave the engagement manager reconciling what was sold with what can be delivered. The estimating answer on this site covers using actual hours to correct the next estimate.
The legal essay recommends that firms start from their own master service agreement and a standard set of terms that covers scope change mechanics, payment terms, termination and limits of liability, and, where a full agreement is not possible, a hardened engagement letter with a minimum set of clauses. As an inference, the contract often decides the question before you do: a fixed fee usually puts estimate risk on the firm, a time and materials agreement puts more on the client, and a clear change clause makes going back routine rather than awkward. The client contract terms answer on this site covers what to include. This is not legal advice.
The essay is direct about timing. When commercial commitments collide with delivery reality, the engagement manager is expected to surface the conflict early and elevate it, and it says silence is not professionalism but risk deferral. It describes AI tracking burn, forecasting contribution margin and flagging erosion early, modeling the economic impact of scope changes, and detecting timeline risk before the client sees it.
As an inference, the question of whether to go back is easiest at 70 percent of budget, when there are still choices, and hardest at 120 percent, when the only choices are an awkward invoice or a loss. The escalation answer on this site covers raising issues internally.
The essay assigns the engagement manager the pricing and contracting of scope adjustments and change orders for the current engagement, with escalation through the delivery manager when commercial commitments collide with delivery reality. It says founders get pulled back into delivery when engagement economics surprise leadership after the fact. As an inference, decide in advance who may approve absorbing an overrun and up to what size, so the decision is made quickly and by someone who can see the margin.
The essay scores the engagement manager on contribution margin first, but adds client satisfaction measured as whether the client is more confident in the firm at the end than at the start, and client lifetime value measured as total contribution margin from first engagement to last. It calls a well-run engagement an annuity protected and a poorly run one an annuity damaged.
As an inference, absorbing a modest overrun for a client with a long future can be a sound investment, and charging for every hour can cost more than it recovers. But absorbing large overruns repeatedly for the same client is a pricing problem, not generosity.
Collective 54 publishes no threshold at which to go back to a client, no change order form and no rule on splitting costs. The published positions are the judgments of when to absorb, push back and protect margin versus relationship, scope creep as an economic decision, the options of reducing scope, extending time or increasing budget, acting early without drama, the client role in each service, activation, the sources of margin leakage, engagement problems as design problems, contract terms and scope change mechanics, early escalation, AI tracking burn and margin, and contribution margin, client confidence and lifetime value as measures.
If the overrun is large enough to make the engagement unprofitable whatever the cause, as an inference, talk to the client about scope now rather than finishing at a loss.
If this is a first engagement with a new client, absorbing a small overrun may buy trust worth more than the hours.
And if it happens on most engagements, the estimate or the offer is wrong; fix that first.
Name the cause first. If the client changed the work or did not do their part, go back to them early, with options to reduce scope, extend the timeline or increase the budget, as the engagement manager essay describes, and without drama or apology. If the estimate or delivery was yours, absorb it, tell the client you are doing so, and fix the estimate. Check what the contract says before the conversation. Watch burn and margin continuously so you raise the issue while there are still choices, and weigh the decision against the whole relationship, not just this engagement.
As an inference from the engagement manager essay, charge when the client changed the scope or did not do their part, and absorb it when the estimate or delivery was yours, telling the client either way.
The engagement manager essay says to raise it early, frame the tradeoff clearly, and offer options such as reducing scope, extending the timeline or increasing the budget, without drama or apology.
As an inference, absorb overruns caused by your own estimate or delivery, and small ones that protect a valuable relationship, but record them and tell the client.
The engagement manager essay describes AI tracking burn, forecasting contribution margin and flagging erosion early, so the conversation happens before the overrun is final.
Sources: Greg Alexander, The AI-Native Boutique Firm (Advantage Books, January 2027), specifically The AI Engagement Manager for the judgments of when to escalate, absorb friction, push back and protect margin versus relationship, scope creep treated as client service, change orders with options to reduce scope, extend the timeline or increase the budget, acting early without drama or apology, margin leakage after the deal closes, early escalation and silence as risk deferral, AI tracking burn, forecasting contribution margin and modeling scope changes, contribution margin, client confidence and client lifetime value as measures, engagement manager authority over change orders, escalation through the delivery manager, founders pulled back by margin surprises, and engagement problems as design problems; The AI Service Design Manager for the role the client must play; The AI Account Executive for activation; The AI Legal Manager for the firm master service agreement, scope change mechanics in the redline playbook and the hardened engagement letter. Related Collective 54 answers on this site: how do I manage scope changes without letting them blow the budget; how do I clearly scope and estimate hours before starting an engagement; what is our process for escalating client issues internally; what terms should we spell out clearly in our client contracts; how do I write proposals and scope engagements so I get paid. Note on scope: Collective 54 is not a law firm and publishes no threshold, form or cost-sharing rule. The four causes, the default for each, telling the client about absorbed overruns, how fee type shifts estimate risk, the 70 and 120 percent illustration, weighing the decision against lifetime value, and the flips are inferences used here to organize the source material rather than published Collective 54 positions.
Collective 54 is the private community for founders and executives of boutique professional services firms between $5M and $50M in revenue. Members work these answers against their own numbers.