Section 1
Key takeaways
• A comp plan is a behavior instrument, not a payroll formula: reps optimize for whatever metric carries the biggest payout, including its destructive version. • Simplicity is a design constraint, not a nicety, Xactly Insights data shows plans built on three components drive sales performance best ; past four variables, small-team reps disengage. • The most expensive comp mistakes aren't about the rate. They're about aiming the incentive at bookings instead of retained revenue, or setting quotas no method can defend. • Over half of companies over-assign quota by 20–30% , and 87% of sales leaders set targets with no structured method, so most "missed quotas" are design failures, not effort failures. • Run a Cobra Test before launch: for every lever, ask what a rep would do if they maximized only that number. If the honest answer damages the business, you found your backfire.
Section 2
Why comp is the highest-leverage number you set
Start with the stakes. Over $1 trillion is spent on sales incentives every year in the United States, commissions, bonuses, and sales contests . That figure belongs to the enterprise giants, but the structural lesson scales all the way down to a founder paying two reps out of a spreadsheet: incentives are the single largest behavioral lever in a sales organization, and they get pulled harder than almost any other. For a small service business, the proportion is even more extreme. As Scott Edinger and Lisa Earle McLeod put it in Harvard Business Review, "in most B2B companies, compensation represents the largest line item in the sales budget" . When your biggest spend is also your strongest behavioral signal, sloppiness isn't a rounding error. It's a strategy, just not the one you intended. Here's the part founders underweight: the people receiving the plan are professionally motivated to find its edges. A good salesperson reads a comp plan the way a tax accountant reads the code, not for what you meant, but for what it literally says. If the plan literally says "you earn the most by booking the largest contract this quarter," a rep will book the largest contract this quarter, even if that means a discount that guts your margin or a client who was never a fit. They're not cheating. They're responding rationally to the instrument you handed them. That's why the rate is a distraction. A 10% commission on the right outcome builds a business. A 10% commission on the wrong outcome dismantles one at exactly the same cost. Before you tune the number, you have to aim it.
Section 3
The Cobra Effect: when a well-meaning incentive makes things worse
The cleanest name for what goes wrong comes from Forma.ai, which frames bad sales incentives through the "Cobra Effect" . The origin story: colonial-era Delhi had a cobra problem, so the government offered a bounty for dead cobras. Enterprising locals started farming cobras to collect the bounty. When the government caught on and scrapped the program, the farmers released their now-worthless snakes, leaving the city with more cobras than before. A well-intended incentive didn't just fail. It worsened the exact problem it was meant to fix. Sales comp is full of cobra farms. Forma.ai names several traps that hit small teams especially hard : Per-product quotas. Attach a separate incentive to each product line and reps will abandon a line the moment they hit its quota, redirecting effort to whatever pays next, even if the customer in front of them needed the first product. You wanted balanced coverage; you bred quota-gaming. Commission caps. Cap earnings and your top reps stop selling once they near the ceiling. Worse, they hoard deals, deliberately delaying a close so it lands in next period when the counter resets. On a six-person team, two reps sandbagging the back half of a quarter can flatten your forecast. CSAT-tied pay. Tie commission to customer-satisfaction scores (CSAT) and reps optimize for the survey, not the sale, coaching customers toward high ratings, avoiding the harder conversations that actually close and expand accounts. Team-pooled incentives. Pool the bonus across the team and your low performers coast on your closers' backs. The free-rider problem isn't theoretical on a small team; it's named and visible, and it corrodes the morale of exactly the people you can't afford to lose. Notice the pattern. Every one of these started as a reasonable idea. "Reward satisfaction." "Encourage teamwork." "Push the new product line." The cobra isn't born from malice or stupidity. It's born from optimizing a single metric in isolation, which is precisely what a salesperson paid against that metric will do. This is the same failure mode that shows up upstream in the funnel, where a lead-gen incentive aimed at raw volume floods the pipeline with unqualified prospects. The mechanism is identical: pay for the proxy, get the proxy, lose the outcome.
Section 4
What structure should a small team actually use?
Once you've accepted that aim beats rate, structure gets simpler, not harder. The market has already converged on a default. Nearly 50% of businesses offer a base salary plus commission, while only 25% offer a base plus a bonus . For most small service teams, base-plus-commission is the right starting point, and the reason is behavioral, not conventional: a meaningful base buys you the right to demand the behaviors that don't show up in a single quarter's bookings, pipeline hygiene, honest forecasting, qualifying out bad-fit deals, while the commission keeps the hunger alive. The temptation, especially for an analytically-minded founder, is to get clever from there. Add a kicker for new logos. A spiff for the strategic product. A multiplier for multi-year terms. A modifier for CSAT. Each addition feels like precision. Collectively, they're a trap. Xactly Insights data is blunt on this: incentive plans with three components are the most effective at driving sales performance . Three. Not seven. There's a cognitive reason, a rep who can't hold the whole plan in their head while talking to a prospect can't be steered by it. Past three or four variables, the plan stops directing behavior and starts producing confusion, and confused reps default to the one behavior they always understand: close whatever's easiest, fastest, biggest. You didn't add precision. You added noise, and noise reverts to the lowest-effort outcome. So the structural recommendation for a small team is deliberately spare: 1. A stable base large enough to buy non-quarter behaviors. 2. One primary commission tied to the outcome you actually want. 3. One guardrail or accelerator that closes the single biggest loophole. That's the whole plan. The discipline isn't in what you add. It's in what you refuse to add.
Section 5
Pay for the outcome, not the proxy
The most consequential decision in the entire plan is what that primary commission attaches to. Most founders default to booked revenue, the contract value at signature, because it's the easiest number to measure and the most satisfying to celebrate. It's also, for a lot of service businesses, the wrong target. Booked revenue rewards the signature and nothing after it. If your business lives or dies on retention, and most service businesses do, then paying full commission at signing trains reps to optimize for getting the yes, not for getting the right yes. They'll discount to close. They'll oversell scope to close. They'll sign clients your delivery team will spend six months apologizing to. Every one of those behaviors maximizes booked revenue and damages the business, which is the textbook definition of a cobra. The fix is to attach the primary commission to retained or collected revenue: revenue that survives onboarding, clears the refund window, and actually lands in the bank. Concretely, on a small team, that can mean paying a portion of commission at signature and the remainder after the client hits a retention milestone (say, the 90-day mark or the second invoice). The rep still gets paid well for closing. But now they're paid more for closing clients who stay, which is the behavior you actually want and the one a booked-revenue plan silently punishes. This is also where comp design has to talk to the rest of the system. A retention-weighted commission only works if the demo and qualification stage is honest about fit instead of optimizing for the yes, and if your follow-up and onboarding are systematized enough that a well-qualified client actually survives to the milestone. Comp can reward the right behavior, but it can't manufacture an outcome the rest of the operation makes impossible. If you want to see how the closing and retention pieces connect end to end, the ConvertOS playbook maps the handoffs.
Section 6
Stop setting quotas you've already decided most reps will miss
If the commission tells reps what to optimize, the quota tells them how high to reach, and quota-setting is where small teams quietly self-sabotage. The data is unsettling. According to ICONIQ Growth research cited by Everstage, 58% of companies over-assign quotas, typically by 20–30% . And per The Sales Collective, 87% of sales leaders set quota targets without a structured method, they pick a number that "feels right," or back into it from a revenue goal they need to hit, with no defensible model underneath. Stack those two facts together and a grim picture emerges: most quotas are inflated and arbitrary. That combination is corrosive in a specific way. A quota a rep believes is unreachable stops functioning as a target and starts functioning as a verdict. People don't sprint toward a wall they're sure they'll hit. They disengage, or they game, sandbagging deals into next period, cherry-picking easy wins, abandoning the long-cycle accounts that don't fit inside an impossible monthly number. An over-assigned quota is itself a cobra: aimed at maximum effort, it manufactures the opposite. For a small team, the prescription is to set quotas you can defend with a sentence. Build the number up from real capacity, average deal size times realistic win rate times the activity a rep can actually sustain, not down from the revenue you wish you had. If you can't explain to a rep why the number is what it is, in plain English, you're in the 87% setting targets by feel, and your people know it. A defensible quota that most reps can hit with strong effort beats an aspirational one that most reps will rationally give up on. The same logic, capacity in, target out, is the kind of incentive-design thinking the Growth Reader unpacks in depth before you commit a number to a plan.
Section 7
Comp plans are not "set and forget"
There's one more mistake that's invisible because it's an absence. You write a plan, it works for a quarter, and then you leave it alone for two years because changing it feels disruptive. Meanwhile the business shifts, new products, new pricing, a move from "grow at all costs" toward predictable revenue, and the plan that once aimed reps correctly is now aimed at last year's priorities. The benchmark says don't do that. Per Xactly's 2021 State of Global Enterprise Sales Performance survey, 75% of revenue leaders made changes to their sales plans over the past year . Revising comp annually isn't churn or indecision; it's maintenance. The companies treating comp as a living instrument are the majority, and they're the ones whose plans still point where the business is going rather than where it's been. For a small team, "review" doesn't require a consulting engagement. It requires one honest annual sitting where you ask three questions: Did anyone game the plan this year, and how? Did the primary commission still point at the outcome we actually wanted? Has anything changed in the business that the plan hasn't caught up to? Those three answers will tell you what to adjust, and crucially, they let you catch a cobra after one breeding season instead of three.
Section 8
The BGA framework: The Cobra Test + the Three-Lever Rule
Here's how to put all of it into a plan you can launch, in order. 1. Name the one outcome you're actually buying. Before any numbers, write a single sentence: "This plan exists to produce more ______." Not "more sales", too vague. "More retained client revenue from good-fit accounts." "More expansion within existing accounts." Whatever it is, it becomes the thing your primary commission attaches to. If you can't name it in one sentence, you're not ready to set a rate. 2. Build exactly three levers, no more. A stable base sized to buy the behaviors that don't show up in one quarter. One primary commission tied to the outcome from Step 1, and weight it toward retained or collected revenue, not just bookings, with a portion held until a retention milestone. One guardrail or accelerator that closes your single biggest loophole. Three components, because three is what Xactly Insights data shows drives performance best , and because a rep has to hold the whole plan in their head to be steered by it. 3. Run the Cobra Test on every lever. For each of the three components, ask out loud: "If a rep maximized only this number and ignored everything else, what damage could they do?" If the primary commission is on bookings, the honest answer is "discount and oversell to close." That's a cobra, go back to Step 2 and re-aim it. Run the test on the guardrail too; guardrails breed cobras as easily as incentives (a CSAT modifier breeds survey-gaming). You're not done until every lever's worst-case answer is something you can live with. 4. Defend the quota in one sentence. Build the number up from capacity, deal size times realistic win rate times sustainable activity, not down from a revenue wish. If you can't explain why the number is the number, in plain English a rep would accept, you're in the 87% setting targets by feel and the 58% over-assigning . Lower it until it's defensible. A target most reps can hit with strong effort beats one most will rationally abandon. 5. Put a review date in the calendar before you launch. Comp is a living instrument; 75% of revenue leaders revise annually . Schedule the one honest annual sitting, Did anyone game it? Did the commission still point at the right outcome? Has the business outgrown the plan?, so you catch the cobra after one season, not three. Run those five steps and the rate genuinely becomes the easy part. You can benchmark it, match it to your market, tune it, and because it's aimed correctly, every dollar you pay pulls the business the direction you wanted.
Section 9
You're running the Cobra Test + Three-Lever Rule right when…
You're running it right when you can hand a new rep the plan and they can explain it back to you in under a minute, base, one commission on retained revenue, one guardrail, without checking notes. When you've actually said the embarrassing sentence out loud for each lever ("if they maxed only this, they'd discount everything") and re-aimed anything that produced a cobra. When your quota survives a skeptical rep asking "why that number?" with an answer better than "because that's what we need to hit." When the largest dollar in the plan is attached to revenue that stays, not revenue that merely signs. And when there's a date on the calendar to revisit all of it, because you've accepted that a comp plan is a behavior instrument that drifts, not a formula you set once and forget. If a rep maximizing their own pay would, as a side effect, build exactly the business you want, you've stopped farming cobras.