Founders ask Collective 54 this 13 times in our records. It usually comes up at the point where informal pay decisions start colliding with each other, or just after someone resigns for more money.
Most compensation arguments happen because the number is being treated as an opinion. It is not.
Greg Alexander method is deliberately unromantic. Determine the role. Calculate the going rate in the market for that role. Pay at the midpoint. Benchmark salary data is widely available, so pick a source and use it. The reason for the midpoint is that a boutique is not a start-up and not a large market leader; it sits between them. You can adjust up or down, but the main point is to pay the market rate.
His framing of why is blunt and worth repeating internally: labor is a commodity priced in an open market, with buyers and sellers, and the equilibrium point between them is the price. If the person quit, that is what they would fetch. If you recruited a replacement, that is what you would pay. Fixing the number to an external benchmark takes the subjectivity out of the discussion, which is the point.
The practical effect in a small firm is larger than it looks. Once salary is a market question, the conversation with a good performer stops being about whether you value them and becomes about which role they are in and what that role is worth. Those are answerable.
Founders often treat below-market pay as a cost saving with a soft cost. The evidence runs the other way.
Alexander describes diligence on a site selection boutique running 40 percent annual turnover. His team contacted former employees to find out why they had left. Role corruption came up, and so did performance reviews that functioned as a compliance exercise. But the plainest finding was this: almost everyone who left the firm landed a job with better pay. The owner was underpaying and the employees knew it. The firm was not invested in.
Turnover also compounds in ways the payroll line does not show. Every departure resets the productivity clock, managers get pulled into rehiring instead of leading, and knowledge walks out. In a small firm, losing one person is an operational event rather than a statistic. And it is contagious: when a respected person leaves, everyone remaining reassesses, which is how a single exit becomes a wave.
It also becomes a diligence question. Buyers ask whether turnover is 15 percent or lower, whether average tenure exceeds five years, whether promotions are filled internally, and whether you are paying your employees what they are worth. Assume every former employee will be contacted.
Salary is solvable with a benchmark. Bonuses are not, and the choice you make there shapes behavior for years.
Alexander frames the underlying question as contribution to wealth creation, since wealth in a boutique is created in only two ways: increasing EBITDA, and increasing the multiple applied to it. A firm at 5 million dollars of EBITDA worth ten times is worth 50 million; a partner who adds 1 million of EBITDA created 10 million of wealth, and one whose work moves the multiple from ten to eleven created 5 million. How much of that should they get is the real question.
There are three systems, each with a cost.
Seniority. Increases are formulaic and tied to years of service. It is easy to administer and understand, it focuses on the long term, it avoids arguments and it smooths year-to-year swings. Its failure mode is young superstars leaving rather than waiting while the long-tenured get paid, and it can keep rewarding past contributions that stopped benefiting the firm.
Performance. Profits are divided by objective measures such as origination, yield and project profitability. It is clear-cut: hit the goals and get paid. Its failure mode is severe in a firm trying to scale, because there is no incentive to build the firm. Referrals are not rewarded, developing staff is not rewarded, and leverage suffers because people drive up their own billability and hoard work rather than pushing it down. Pay that prioritizes short-term performance can destroy the ability to scale.
Reconciliation. Profits are distributed after the group reconciles how wealth was created for one another, using a consistent methodology, collected evidence and explained logic. Its advantages are balance between short and long term, few fights because everyone is on the jury, no single king ruling the decision, and a clear signal about what the firm values. Its costs are real too: attribution is genuinely hard, it is time consuming, it requires full disclosure so everyone knows what everyone got, and it strains as the number of partners grows.
The choice is not about fairness in the abstract. It is about which behavior you can afford to encourage.
This is the mistake Alexander is most direct about, because it was his.
As SBI moved from start-up to boutique he did not change the partner compensation system. Partners were paid a very generous salary that was not market based, with the first group earning roughly twice what they could fetch in the open market. Bonuses were paid as distributions, and distributions were tied to equity stake, so a partner owning 25 percent received 25 percent of distributions.
The rigidity is the lesson. Changing equity stakes is much harder than changing bonus payouts. As a new group of partners began to contribute more than the legacy group, the dollars should have shifted and could not. The contribution gap widened, relationships fractured, and he describes friends becoming enemies. He calls it a failure of leadership on his part, and says it cost millions.
The design principle that falls out of it is simple. Salary answers what the role is worth. Bonus answers what happened this year and can be recalculated annually. Equity answers who owns the firm, and it should be changed rarely and deliberately. Collapsing any two of the three creates a mechanism you cannot adjust when the facts change.
Compensation is constrained by yield, and a compensation problem is often a pricing problem wearing a disguise.
Yield is average fee per hour multiplied by average utilization. The benchmark fee levels Alexander uses run above 750 dollars for senior staff, above 500 for midlevel and above 250 for junior, against an average above 400. If your fee levels sit well below that, no compensation design will let you pay at market and still make money.
The lever is not squeezing utilization, which most firms past start-up have already optimized. It is fee level, and the route there is specialization across industry, function, segment, problem and geography. Labor is also the biggest expense in a boutique, which is why the ability to match supply and demand matters as much as the rate itself.
So before redesigning a compensation plan, check whether the plan is the constraint or whether the business model is.
In Era 1, pay decisions were made under pressure and without forward visibility, alongside hiring that surged during overload and froze months later. Era 2 digitized the paperwork without changing the outcomes, and the presence of data was mistaken for insight.
What moves in Era 3 is the quality of the inputs. Time to productivity can be modeled by role and skill set, which tells you what a person is actually worth to the firm and when. Attrition risk can surface early, which changes a retention conversation from reactive to planned. Promotion readiness accumulates as a signal rather than arriving as an annual judgment. Capacity can be forecast against projected demand, which is what makes a hiring and pay plan something other than guesswork.
What does not move is the decision. Judgment, tradeoffs, cultural reinforcement, coaching and accountability stay human. AI does not set the number, and a pay conversation that feels generated will land exactly as well as a performance review that feels generated.
Below roughly ten people, a formal band structure is overhead. What you still need is the market benchmark, so that each number can be explained without reference to how you feel about the person.
If you genuinely cannot afford market rate, compensation design is not your problem. Fee level, utilization mix or leverage is, and paying below market while you work that out is a decision with a known cost rather than a saving.
And equity is not a substitute for cash with employees who need cash. Alexander early partners took equity instead of cash and that created a debt he then paid in perpetuity because he never defined when it was settled. If you offer equity in place of salary, define what it is replacing and for how long.
Set salaries from external benchmarks and pay at the midpoint, because labor is priced in an open market and a boutique sits between a start-up and a market leader. That removes most of the argument. Do the real design work on the bonus, choosing consciously between seniority, which is easy to run but drives young performers out, measured performance, which is clear-cut but destroys leverage because people hoard billings, and reconciliation against contribution to wealth creation, which balances best but demands disclosure and time. Keep salary, bonus and equity separate, since equity is rigid and the hardest thing to unwind. Treat underpaying as a cost rather than a saving: your people know the market, turnover is contagious, and buyers will ask whether you paid fairly and will contact your former employees. Check the economics before redesigning the plan, because if your fee levels do not support market pay the constraint is pricing and specialization rather than compensation. In Era 3 better inputs are available, but the number is still a human decision.
Define the role, find the going market rate for it, and pay at the midpoint of the benchmark. A boutique is neither a start-up nor a large market leader, so the midpoint is the right anchor, and you can adjust from there. The reason to fix the number externally is that labor is a commodity priced in an open market: it is what the person would fetch if they left and what you would pay to replace them. Anchoring to a benchmark takes the subjectivity out of a conversation that otherwise turns into a referendum on how much you value someone.
Three, and each has a cost. A seniority system ties increases to years of service: easy to administer, long-term in focus, but it pushes young high performers out and can keep rewarding contributions that stopped mattering. A performance system divides profits by objective measures such as origination, yield and project profitability: clear-cut, but it removes any incentive to build the firm, because referrals and developing staff are unrewarded and partners hoard billings rather than pushing work down. A reconciliation system distributes profits after the group agrees how wealth was created: it balances short and long term and reduces fights, but attribution is hard, it takes time, and it requires full disclosure.
Be careful, because equity is the rigid instrument. Changing equity stakes is much harder than changing bonus payouts, and Greg Alexander describes tying distributions to equity stake at SBI as a mistake that cost millions and fractured partnerships, because the money could not shift as newer partners began contributing more. Salary should answer what the role is worth, bonus should answer what happened this year and be recalculated annually, and equity should answer who owns the firm and change rarely. If you do offer equity in place of cash, define what it is replacing and for how long.
More than the saving. In diligence on a firm running 40 percent annual turnover, almost everyone who had left landed a better-paying job and the employees knew they were underpaid. Every departure resets the productivity clock, pulls managers into rehiring and lets knowledge walk out, and in a small firm turnover is contagious, because one respected person leaving prompts everyone else to reassess. It is also a diligence question: buyers ask whether turnover is 15 percent or lower, whether tenure exceeds five years, whether promotions are internal and whether you pay people what they are worth, and they contact former employees.
Sources: Greg Alexander, The Boutique: How to Start, Scale, and Sell a Professional Services Firm (Advantage, 2020), chapter 23 on partner pay, market-based salaries at the midpoint, wealth creation through EBITDA and the multiple, the seniority, performance and reconciliation bonus systems and their tradeoffs, and Greg account of the compensation mistakes he made at SBI, chapter 35 on employee loyalty, the site selection boutique at 40 percent turnover, underpayment, the 15 percent turnover benchmark, tenure and internally filled promotions, chapter 21 on labor as the largest expense and matching supply to demand, and chapter 14 on yield and the benchmark fee levels by seniority. Greg Alexander, The AI-Native Boutique Firm (Advantage Books, January 2027), specifically The AI HR Manager for the Era 1 and Era 2 account of pay and hiring decisions made without forward visibility, and for time to productivity modeling, attrition risk detection, promotion readiness and capacity forecasting as continuous signals in Era 3, with judgment, tradeoffs and accountability remaining human. The SBI and Capital 54 accounts are Greg own experience.
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.