Founders ask Collective 54 this once in our records, and not in 2026. The outcome-based pricing and not delivering as promised answers on this site cover pricing on results and what happens after a miss; this page covers whether to guarantee the work at all.
The competitor chapter of the 2020 book names five competitors for every boutique and says about 20 percent of the time the competitor is another boutique, often because the client has budget constraints. Its remedy is to guarantee the work. It says boutiques have limited resources, which makes them risk averse, and that a guarantee frightens them because they may not get paid. By guaranteeing your work, it says, you separate yourself from boutique competitors. Its checklist asks whether you are confident enough to guarantee your work.
The referral essay in the newer book explains the buyer side: professional services are leap-of-faith purchases with no demo and no return policy, and the buyer must believe before evidence exists. As an inference, a guarantee is a return policy for judgment. It moves risk from the buyer to the firm, and on a large contract that is exactly the risk a buyer has to explain to their own leadership.
The account executive essay says a buyer must be able to justify the decision internally before an opportunity advances. As an inference, a guarantee helps with that justification, but the larger the contract, the larger the exposure if you misjudge it. A full refund on a year-long engagement could erase the profit of several others. The answer is not to avoid guarantees on large work but to design them deliberately.
The pricing essay lists outcome-based and risk-sharing models among the pricing approaches that introduced alignment between firm incentives and client results, particularly in high-confidence engagements. The outcome-based pricing answer on this site covers choosing outcomes the firm can influence and measure. As an inference, guarantee something within your control and observable: that a deliverable meets an agreed standard, that a milestone is reached by a date, or that a defined result is achieved under stated conditions. Avoid guaranteeing outcomes that depend mostly on the client, the market or other vendors.
The service design essay in the newer book says services should be defined by outcomes, with explicit scope boundaries and a clear statement of the role the client must play. The delivery professional essay describes acceptance criteria and a definition of done as the answer to uncertainty about whether work is finished. As an inference, a guarantee without an agreed standard is an argument waiting to happen. Write down what success means, how it will be measured, and who decides, before the work starts. The clear deliverables and rejected deliverables answers on this site cover how.
As an inference, most guarantees fail not because the firm underperformed but because the client did not provide data, access, decisions or people on time. Make the guarantee conditional on the client role you described at the sale: timely input, named decision makers, and adoption of what you deliver. This is not a loophole. It is what makes the guarantee fair and keeps both sides focused on the result.
The legal essay in the newer book lists limitation of liability and indemnities among the terms a firm must control, and warns against broad or uncapped indemnities accepted by default. As an inference, the remedy should be proportionate: first, additional work at no charge until the standard is met; then, if that fails, a refund of the fees for the affected phase rather than the whole contract. Cap the total at the fees paid. Collective 54 gives no legal advice, so have your attorney draft the guarantee and check how it interacts with the rest of the contract.
As an inference, a large contract is easier to guarantee in phases. Guarantee the first phase fully, so the client can test you at low risk, and let later phases carry outcome-based or milestone terms. The fee quality chapter of the 2020 book says buyers of firms value services that build on one another, from assessment to solution to implementation; a guaranteed first phase is a natural entry to that sequence.
The competitor chapter says that about 5 percent of the time a boutique competes with the market leaders, and that these deals are the year makers. Its five steps for beating them are to establish credibility, deliver a top-quality proposal, show you can finish much faster, offer the same quality for about 25 percent less without discounting so far that you look cheap, and offer an easier experience than a large firm that enters a client like a tornado. As an inference, a guarantee strengthens the credibility step on large contracts in particular, because it answers the question a buyer has about choosing a smaller firm: what happens if this goes wrong. The prospect financials answer on this site makes the same point.
The pricing essay says pricing must be governed rather than set once, and lists controlling discounting and exceptions, aligning delivery with pricing assumptions and connecting pricing behavior to realized margins among the work involved. As an inference, decide in advance which offers carry a guarantee and on what terms, rather than offering one ad hoc to close a deal. Track how often guarantees are invoked and what they cost. If they are invoked often, the problem is delivery or scoping, not the guarantee.
The not delivering as promised answer on this site recommends honoring a guarantee promptly and without argument, because a guarantee that has to be fought for is worse than none. As an inference, the reputational value of a guarantee comes mostly from how you behave on the rare occasions it is called. A client who sees you keep your word, even at a cost, is likely to become an advocate.
Collective 54 publishes no guarantee wording, refund policy or cap and gives no legal advice. The published positions are guaranteeing the work to stand apart from risk-averse boutiques, the confidence test, leap-of-faith purchases, buyers justifying decisions internally, outcome-based and risk-sharing models for high-confidence engagements, governed pricing exceptions, services defined by outcomes with client roles, acceptance criteria and a definition of done, controlling liability and indemnities, and services that build on one another.
If the result depends mostly on things you cannot control, as an inference, guarantee your process and responsiveness instead of the outcome.
If your work is new or unproven, gather evidence on smaller engagements before guaranteeing a large one.
And if the client is a large organization with its own contract terms, its legal team may reshape any guarantee; negotiate from your own wording with counsel.
A guarantee can set you apart; the 2020 book recommends guaranteeing the work because risk-averse boutiques will not. On a large contract, make it narrow and deliberate: guarantee an outcome you control, defined and measured in advance, conditional on the client doing its part, with a proportionate remedy capped at the fees for the affected phase. Phase large work so the first phase carries the guarantee. Govern guarantees as pricing exceptions, honor them without argument, and have your attorney draft the terms.
The 2020 book recommends guaranteeing the work to separate a firm from risk-averse boutique competitors, and asks whether you are confident enough to do so. As an inference, guarantee outcomes you control, defined in advance.
As an inference, define the standard in advance, make it conditional on the client role, offer rework before refund, cap the remedy at the fees for the affected phase, and have counsel draft it.
As an inference, no. The pricing essay lists outcome-based and risk-sharing models as pricing approaches; a guarantee is a remedy if an agreed standard is not met. Both suit high-confidence engagements.
The not delivering as promised answer recommends honoring it promptly and without argument. As an inference, frequent claims point to delivery or scoping problems.
Sources: Greg Alexander, The Boutique: How to Start, Scale, and Sell a Professional Services Firm (Advantage, 2020), chapter 3 for the five competitors, other boutiques about 20 percent of the time, the five steps for beating market leaders, guaranteeing the work, boutiques as risk averse and frightened of guarantees, and the confidence self-check; chapter 32 for services that build on one another from assessment to solution to implementation. Greg Alexander, The AI-Native Boutique Firm (Advantage Books, January 2027), specifically The AI Pricing Manager for outcome-based and risk-sharing models in high-confidence engagements and pricing governance including controlling exceptions and connecting pricing to realized margins; The AI Referral Generator for leap-of-faith purchases; The AI Account Executive for buyers justifying decisions internally; The AI Service Design Manager for services defined by outcomes with scope boundaries and client roles; The AI Delivery Professional for acceptance criteria and a definition of done; The AI Legal Manager for limitation of liability and indemnities. Related Collective 54 answers on this site: how do I design, measure, and roll out outcome-based pricing; what happens if we do not deliver as promised; how do I define clear deliverables so clients know what they are buying; how do I avoid clients rejecting deliverables we did not confirm with them upfront; what terms should we spell out clearly in our client contracts; what do I say if a prospect asks about our financials. Note on scope: Collective 54 publishes no guarantee wording, refund policy or cap and gives no legal advice. The guarantee as a return policy for judgment, guaranteeing what you control, conditions tied to the client role, rework before refund, capping at phase fees, phasing large work, governing guarantees as exceptions, 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.