Founders ask Collective 54 this 8 times in our records, 2 of them in 2026. Almost always after the first retainer turned into unpaid unlimited access.
The reason retainers get mispriced is that firms price them as a discounted block of hours, which is not what the client wanted.
Clients on retainer are paying for institutional knowledge, fast access to a trusted advisor, strategic responsiveness and availability across time rather than tasks. Revenue decouples from hours worked, and clients pay for continuity rather than capacity. That distinction is the whole pricing argument. Hours are what it costs you. Continuity is what they are buying.
This is why a retainer priced as twenty hours at a discount is worse than useless. It teaches the client to count hours, which is the one frame in which your value shrinks every time you get faster.
Two benefits, and both matter more than the fee level.
The first is cash. A retainer means getting paid in advance, and paid in advance is a marker of fee quality that buyers look for directly. Firms with aging receivables have poor fee quality. Firms paid up front can fund growth from free cash flow rather than short-term debt, which is exactly what financial buyers want to see.
The second is predictability. Monthly recurring revenue stabilizes cash flow and makes EBITDA more predictable, and recurring revenue carries a higher multiple than nonrecurring revenue. Retainers also push average contract length past twelve months, which is another thing buyers examine, because a firm selling thirty-day assessments has poor fee quality and a firm on twelve-month agreements does not.
Relationships deepen too, which improves retention and gives you the natural surface for upsell and cross sell.
There are only so many retainers a boutique can handle at one time. That sentence is the constraint that decides how you price.
A retainer reserves capacity whether or not the client uses it, and reserved capacity you cannot resell is real cost. So the floor on a retainer price is not the hours you expect to spend. It is the value of the capacity you are taking off the market, which is a larger number in every firm that is close to full.
This is also the honest limitation of the model. Retainers are still people-delivered, still judgment-heavy, and still constrained by human capacity and availability. They are the final evolution of the people-delivered firm, and they signal maturity rather than modernity. They do not, by themselves, carry a firm into the AI-delivered era.
Four things belong in the agreement, and getting them written is worth more than getting the number exactly right.
Define the advisory lanes. Say specifically what the client has access to, and by implication what they do not. Vagueness here is generous in month one and expensive by month six.
Package into tiers with clear boundaries. Versioning lets the client choose their own price, which speeds up the decision and links price to value in their own mind. Price is what they pay and value is what they get, and the two are never the same. A tier structure makes the client tell you which one they are buying.
Set governance expectations explicitly. Response times. Meeting limits. Who can call. What constitutes a new project rather than a covered request. These are not bureaucratic; they are what makes the fee defensible when usage spikes.
Build in an annual increase. If it is not in the agreement, raising it later becomes a negotiation instead of a term.
One note on the number itself. Perception is reality in pricing, and a retainer priced too low reads as low quality rather than as good value, particularly in a model where the client cannot see the work continuously. The fee is part of how the client understands what they have bought, so set it against the position you want to hold rather than against what you fear they will accept.
The dominant failure mode is scope drift disguised as good service. An exception becomes a norm, customization creeps in, and the boundary between what is included and what is extra stops being visible to anyone. Most pricing breakdowns happen this way rather than through the headline price, and the only defense is watching the frequency of exceptions rather than their individual merits.
A price that is not enforced is not a price. If deviations from the agreed structure are not visible, discounting and scope expansion erode the margin quietly, and nobody can say when it started.
The second failure is the one with an AI twist. If part of the work behind the retainer is now performed by AI and delivery got materially faster, the pricing assumption behind the fee has aged. That efficiency will leak, passed to the client unintentionally or absorbed by expanding scope, unless someone decides deliberately what happens to it. Faster delivery inside a fixed monthly fee is a margin gain only if the fee holds.
Do not run a single revenue source. Boutiques have nine to choose from: hourly billings, retainers, fixed bids, performance-based contracts, memberships, licensing, subscriptions, events and royalties. The firm Greg Alexander built ended up roughly a third retainers, a third fixed bids and a third performance-based fees.
Aim the balance at fee quality rather than at retainers specifically. The rough rule of thumb is sixty percent of fees from existing clients and forty percent from new, since firms addicted to new-client revenue burn cash to generate it, and firms over-indexed to existing clients forget how to hunt and wake up unable to.
If your work is genuinely episodic, with long gaps and no need for standing access, a retainer is a worse fit than a fixed bid. Do not convert a project business to retainers because recurring revenue sounds better at exit; a retainer nobody uses gets cancelled, and cancelled recurring revenue is worse for your numbers than a healthy project business.
If you are running close to capacity, be careful. A retainer that reserves time you could sell at project rates is a margin decision, not just a stability decision, and the answer depends on what the alternative work pays.
And if a client is pushing you toward a retainer specifically to lower their effective rate, the conversation is about price, not structure. Have that conversation directly.
Price the access, not the hours. A retainer is payment in advance to secure your services when needed, so what the client is buying is institutional knowledge, fast access, strategic responsiveness and availability across time, which means pricing it as a discounted block of hours both undersells it and teaches the client to count hours. Price against the capacity you are reserving, since there are only so many retainers a firm can carry and reserved capacity you cannot resell is a real cost. Structure it before you price it: define the advisory lanes, package into tiers with clear boundaries so the client picks their own price, write governance terms covering response times and meeting limits, and build in the annual increase. Then watch for the two failure modes, exceptions quietly becoming norms until the boundary disappears, and AI making delivery faster without anyone deciding who keeps the gain. The payoff is cash in advance and predictable monthly revenue, both of which raise fee quality and the multiple, but retainers remain people-delivered and capacity-constrained, so treat them as one third of a revenue mix rather than the destination.
Access and continuity, not hours. A retainer is a client paying in advance to secure your services when needed, and what they value is institutional knowledge, fast access to a trusted advisor, strategic responsiveness and availability across time rather than tasks. Revenue decouples from hours worked. Pricing a retainer as a discounted block of hours undersells it and teaches the client to count hours, which is the one frame where your value falls every time you get faster. The practical floor is the value of the capacity you are reserving, because there are only so many retainers a boutique can carry at once.
Four things. Defined advisory lanes, so what is covered and what is not is explicit rather than generous by default. Tiers with clear boundaries, which let the client choose their own price and link price to value in their own mind, since price is what they pay and value is what they get. Governance expectations, meaning response times, meeting limits, who can call, and what counts as a new project rather than a covered request. And a built-in annual increase, because a raise that is not a term becomes a negotiation.
Packaging drift. An exception becomes a norm, customization creeps in, and the line between included and extra stops being visible. Most pricing breakdowns happen this way rather than through the headline price, and the defense is tracking how often exceptions occur rather than judging each one on its merits. A price that is not enforced is not a price. There is also an AI version of the problem: if AI has made delivery materially faster, the pricing assumption behind the fee has aged, and that efficiency leaks into unintentional pass-through or expanded scope unless someone decides deliberately who keeps it.
Favorably, on three of the measures buyers use. Getting paid in advance is a marker of high fee quality, because firms with aging receivables look weak and firms funding growth from free cash flow rather than short-term debt look strong. Recurring revenue carries a higher multiple than nonrecurring revenue and makes EBITDA more predictable. And retainers push average contract length past twelve months, which buyers examine directly. The caveat is that retainers stay people-delivered and capacity-constrained, so they signal maturity rather than modernity.
Sources: Greg Alexander, The Boutique: How to Start, Scale, and Sell a Professional Services Firm (Advantage, 2020), chapter 4 for the definition of a retainer as a client paying in advance to secure services when needed, for the benefits of advance payment and predictable cash flow, for the constraint that there are only so many retainers a boutique can handle at one time, for the nine revenue sources of hourly billings, retainers, fixed bids, performance-based contracts, memberships, licensing, subscriptions, events and royalties, for the warning against going to market with a single source of revenue, and for the mix at the author own firm settling at roughly one third retainers, one third fixed bids and one third performance-based fees; chapter 15 for price versioning allowing clients to choose their own price and decide faster, for the distinction that price is what you pay and value is what you get, for perception as reality in pricing and the signal a price level sends about quality, and for building an annual price increase into the system; chapter 32 for fee quality, including recurring revenue commanding more from acquirers than nonrecurring revenue, contracts longer than twelve months, collection in advance versus aging receivables, funding growth from free cash flow rather than short-term debt, and the rough sixty forty split between existing-client and new-client fees. Greg Alexander, AI Pricing Strategy, Collective 54, for retainers as the Exit stage of Era 1, with clients paying for continuity rather than capacity and valuing institutional knowledge, fast access, strategic responsiveness and availability across time, for monthly recurring revenue stabilizing cash flow and making EBITDA more predictable, for the limitation that retainers remain people-delivered, judgment-heavy and capacity-constrained and therefore signal maturity rather than modernity, and for the conversion playbook of defining advisory lanes, packaging into tiers with clear boundaries and setting governance expectations on response times and meeting limits. Greg Alexander, The AI-Native Boutique Firm (Advantage Books, January 2027), specifically The AI Pricing Manager for packaging discipline and the finding that most pricing breakdowns occur through packaging drift rather than headline prices, with alerts needed when exceptions become norms, for the position that a price which is not enforced is not a price, and for delivery and pricing alignment checks that catch delivery speed increases and AI substitution before pricing logic ages.
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.