← Contents08 / 20

Module 8 — Blood Money: Designing the Comp Plan

"Show me the comp plan and I'll show you the company. Everything else — the values poster, the all-hands, the gong — is set dressing. The plan is the only thing on the wall anybody actually reads, and they read it like it's the only honest document in the building. Because it is." — my attorney, who bills at $650 per hour to say things I already know

The comp plan is the most truthful document your company will ever produce, which is precisely, exquisitely, unbearably why everyone wants to make it dishonest. It is a contract written in money that tells every salesperson on your floor — in language so clear it strips away every PowerPoint abstraction, every culture deck aspiration, every all-hands inspirational video featuring outdoor imagery — exactly what you will pay them to do. And they will do that. Precisely that. Nothing adjacent to that. Nothing adjacent-to-adjacent to that. With the cold, efficient, tireless devotion of a heat-seeking missile that has a lease payment, a dental plan, and a kid starting private school in September.

You do not get the behavior you want. You do not get the behavior that lives in your values deck. You do not get the behavior you described at the last all-hands while everyone nodded and someone hit the gong. You get the behavior you pay for, down to the last decimal, and if there is a loophole in that plan — a gap between what you intended and what the document actually says — a rep making $190K OTE with a Patek he can't quite afford and a mortgage he definitely cannot will find that loophole before lunch on the first Monday of the quarter, and he will drive a goddamn freight train through it while you're in a pipeline review talking about "alignment." Comp design is not an HR exercise. It is not a finance checkbox. It is applied behavioral engineering with a payroll deadline, and it is the highest-leverage lever you own. Treat it like one or spend Q4 wondering why your team is selling the exact wrong thing with stunning efficiency.

THE JOB

The comp plan converts strategy into individual human incentive. Its job — its only job, its whole reason for existing — is to make the thing that's good for the company also the thing that makes the rep rich, and to do it so cleanly, so mechanically, that the rep never has to choose between their paycheck and the right call. Every dollar of commission structure is a sentence in a behavioral instruction set. Write it carelessly and you've programmed the precise disaster you'll spend the next four quarters firefighting and explaining to a board that has mysteriously lost patience.

The governing law of this entire module, and honestly of most of your professional life in this business:

RULE No. 8: You get the behavior you pay for. Not the one you wanted. Not the one in the deck. Not the one on the wall. The one in the goddamn plan.

THE PLAYBOOK

Step 1. OTE and pay mix — the foundation, and also the first fight

OTE (On-Target Earnings) = total cash a rep earns at exactly 100% of quota. It splits into base salary (guaranteed, lands whether they close nothing or close everything) plus variable compensation / commission (earned by hitting things). The split between these two numbers is the pay mix, written as base/variable:

AE — closing role: 50/50. Half guaranteed, half at risk. This is the industry default and it is a good one. The risk is real. The upside is real. The alignment is real.
SDR / BDR: 60/40 to 70/30. More base. SDRs don't close revenue; they create the conditions for someone else to close revenue. They deserve stability — and frankly, a 50/50 split on a role that depends heavily on marketing pipeline quality and manager coaching is how you end up with SDRs who are broke and pissed off and booking calendar-filler meetings to hit activity metrics.
AM / Renewals / CSM-with-quota: 60/40 to 80/20. Significantly more base. You do not want a renewals rep negotiating against a customer you need to retain for five years like she's trying to slam shut a new logo deal before the EOQ Slack pops off. That customer will notice. They will leave. That is the opposite of renewal.

The pay mix is a message. Heavy variable says: "high risk, high autonomy, you are the hunter, kill what you eat, the upside is yours if you can grab it." Heavy base says: "stability, retention, relationship, do not do anything insane to make a short-term number." Match the mix to the actual behavioral profile the role requires. Mismatch it and you've sent the wrong instructions to the wrong people and you will get back exactly what you asked for.

Step 2. Accelerators and decelerators — bending the slope

Commission should not be a straight line from zero to forever. A flat rate at every attainment level is the laziest possible comp design and it rewards average performance at the same rate it rewards extraordinary performance, which is the mathematical equivalent of giving your top closer a participation trophy.

Bend the curve.

Accelerators: Above 100% attainment, the commission rate increases. Standard: 1.5× the base rate between 100–125%, stepping to 2× above 125%. Some plans add a third tier above 150% for the truly deranged overperformers who are carrying half the number on their backs. The accelerator exists for one reason: to make your top 20% obscenely motivated to keep closing after they've already hit plan, instead of sandbagging the last two deals of the quarter into January so they start next year ahead. The accelerator is the engine of overperformance. Fund it. Do not cap it out of fear that someone might earn too much. A rep who earned $400K because she closed $3M in a $180K OTE role is the best possible news in your building. Throw a party. Send a fruit basket. She just carried everyone else's miss.

Decelerators and floors: Below some floor — commonly 50–70% of quota — the commission rate drops, or commission doesn't begin until a minimum threshold is crossed. This mechanic deserves respect and restraint. A reasonable commission floor (zero payout below, say, 40% attainment) is defensible business logic. A brutal clawing decelerator structure below 80% is a message to your struggling reps: we think you might be beyond saving and we've built that assumption into the math. They will receive that message. They will update their LinkedIn before you finish composing your feedback.

Step 3. Quota retirement — what counts and how much, which is the actual steering wheel

Quota retirement = the mechanism by which a booking reduces a rep's remaining quota balance. This is where you steer the deal mix with genuine financial precision. You decide, here, at the level of the plan document, what types of deals the company actually fucking wants — so make the decision deliberately and put it in writing.

Examples:

New logo retires at 1.0×, but a multi-year prepaid deal retires at 1.2× → you've just embedded a financial preference for cash and term length directly into the rep's wallet. They will sell more multi-year deals. This is not magic; it is paying people to do the shit you need done.
Services revenue or one-time implementation fees retire at 0.5× or not at all → reps stop bundling low-margin professional services into deals to pad bookings numbers. The padding was real before. It will stop being real the instant the plan makes it financially stupid to continue.
A deal closed at a heavy discount — say, 30%+ off list — retires at a haircut: 0.8× or less → discount discipline, enforced at the source, in the paycheck, where the rep will feel it immediately. (We tear into this in Module 14, and it gets genuinely ugly.)

Most comp plans never touch quota retirement. They leave it flat at 1.0× across every deal type and then waste all their management energy screaming at reps about deal mix — as if screaming were a substitute for just paying people correctly in the first goddamn place. The retirement rules are the steering wheel. Turn the wheel.

Step 4. Draws — recoverable vs. non-recoverable, and why one of these will end in a lawsuit

A draw is guaranteed commission income, usually deployed for ramping reps or for reps transitioning into a new territory or plan, so they can pay their rent while their pipeline builds. Draws sound generous. They are also a landmine if you pick the wrong flavor.

Non-recoverable draw: The company eats it. Rep keeps it regardless of what they close. Use this for ramp — you are investing in a new hire's survival while they learn the product, the territory, and how to spell the customer's name correctly. Do not claw back their food money in month two because they haven't sourced enough pipeline yet. That is monstrous and also it will make them quit.

Recoverable draw: An advance against future commissions. The rep must earn it back; if earned commissions in future months don't cover the draw balance, in theory they owe the company money. This is brutal. It is often legally complicated. It is a fast way to turn a new hire's anxiety into active resentment and then into a resignation email. Use rarely. Label it in screaming capitals in the offer letter. Explain it verbally. Explain it again.

My attorney advises: recoverable draws are a lawsuit with a payroll schedule. Several states make recovering draw balances from a departed employee nearly impossible in practice, and attempting it will cost you more in legal fees than the draw was worth — plus a Glassdoor review that will make your next round of recruiting significantly more expensive. Default to non-recoverable for ramp situations and sleep at night like a person who made a sensible decision.

Step 5. Clawbacks — getting the money back when the deal dies

A clawback reverses commission already paid when a deal collapses after payout: customer churns inside 90 days of close, invoice never pays, deal cancels before any product is delivered. Clawbacks are necessary infrastructure — not optional, not cruel, not a "trust issue." Without them, the rational move for every rep is to close anything that will sign, pocket the commission, and let Customer Success inherit the smoldering wreckage while they're already three calls deep on the next deal. The CS team will learn to hate Sales. They already hate Sales. The clawback is the structural argument against giving them more reasons.

Design them right:

Tie to a specific, clear trigger — non-payment after X days, cancellation inside the clawback window, defined in the plan document, not improvised in a Slack message.
Set a reasonable window — 90–180 days is standard. Much beyond 180 and you're clawing back money from deals that ran their course rather than deals that were always garbage.
Keep the language narrow and precise. A vague, sprawling clawback clause that reps interpret as "the company can take my commission back for any reason they feel like" poisons the floor faster than a bad manager. Reps stop believing any commission is ever truly theirs, and a rep who doesn't trust the plan stops selling with full effort. You've just cut your own payroll expense in the worst possible way.

Step 6. SPIFFs — the short-term cattle prod with a short shelf life

A SPIFF (Sales Performance Incentive Fund — the acronym is an artifact of a more innocent era, before anyone understood what they were doing) is a one-time bonus for a specific behavior inside a tight window: $500 per deal on the new product this month, $2,000 for the first logo landed in the target vertical, $1,000 for every net-new logo closed before the 25th. SPIFFs are espresso shots — sharp, effective, and deeply ruinous as a long-term dietary strategy. They are also, used carelessly, a way to tell your floor that the normal comp plan doesn't fucking work and you've had to supplement it with performance bribery. That message lands. Keep the SPIFFs surgical.

Use them to:

Launch a new product and force reps to learn it by making learning it pay.
Attack a specific segment or geographic market with focused intensity.
Clear a specific pipeline logjam at end-of-quarter without restructuring the whole plan.

Hard rules:

Time-box relentlessly. A SPIFF that never ends is not a SPIFF. It is an undocumented comp plan addendum that reps have now internalized as permanent, and the day you try to remove it you will have a floor-level uprising.
One SPIFF at a time, maximum two. More than that and you've created a complexity problem inside a tool designed for simplicity. Reps will optimize for the SPIFF math and the base plan will drift out of their mental priority list. You'll get SPIFF behavior, not quota behavior.

Step 7. Sample plan structures by role — actual numbers, because vague examples are useless

SDR / BDR

OTE range: $70–90K, pay mix 65/35 base to variable.
Variable earns on SQLs accepted and converted to pipeline — not raw meetings booked, not activities logged, not the sheer volume of LinkedIn connection requests. If you pay on raw meetings booked you will get a calendar full of no-shows, zombie leads, unqualified contacts who had no idea what the product does, and AEs who develop a slow-burning hatred of the SDR team that poisons the floor culture for years. Pay for quality. The SQL qualification criteria must be in writing, the AE must confirm them, and there must be a clear arbitration process for disagreements — or the SDR-AE relationship becomes a two-person warfare unit that sucks the energy out of every pipeline review.
Small kicker on sourced pipeline that converts to closed-won. This aligns the SDR to revenue outcome, not just top-of-funnel vanity. It takes a quarter to settle in and it fundamentally changes how SDRs think about ICP targeting.

AE — Mid-Market

OTE range: $140–180K, pay mix 50/50.
Variable on net-new ARR closed — not bookings, not TCV of multi-year unless you've set retirement accordingly, not contract value with services bundled in.
Accelerators: 1.5× rate from 100–125%, 2× rate above 125%. No fucking cap. The moment you put a cap on the accelerator you've told your best rep that excellence has a ceiling and she should stop at that ceiling and wait. She will not wait. She will find an employer who will pay her more for the same performance and she will be gone before you finish the comp revision memo.
Quota retirement: multi-year prepaid at 1.2×; professional services at 0.5×; discounts above 25% carry a 0.85× haircut.
90-day non-payment clawback, clearly written.
Non-recoverable ramp draw for months 1–3. After month 3, it's on them.

AM / Renewals

OTE range: $120–160K, pay mix 70/30 (this is a retention and relationship role; heavy base is correct).
Variable splits between two separate components that you should not crush into one number:
- Retention / GRR component — paid on keeping the existing book of business. If the customer renews, this piece pays. If the customer churns, this piece does not pay. Clean, direct, and aligns the AM to the thing that actually matters.
- Expansion / upsell component — paid on growing the book via upsell and cross-sell. Runs like a mini-AE commission, often with a lower rate and a more modest target.
Do not, under any circumstances, make the renewal rep's primary income dependent on expansion. You will get a renewal rep who neglects her renewals to chase upsell deals — because upsell pays more — and then loses the renewal because she was too busy chasing the upsell. You have now lost both. That customer is gone. The churn hits your NRR. The board asks why retention is declining. The answer is that you paid for this outcome, specifically, with a comp plan you designed badly. You built a machine that produced exactly this result. Congratulations on the engineering.

Step 8. Complexity — the gaming tax, and why three components is a hard ceiling

RULE No. 8b: If a rep cannot compute their own commission on a cocktail napkin, the plan is already broken, and you're going to find out how broken in the most expensive possible way.

Complexity is the enemy. Full stop, no asterisk, no "well it depends." Every component, every modifier, every quarterly multiplier, every "momentum index" dreamed up by some revenue consultant who bills by the component is simultaneously a place to game the plan and a reason reps stop trusting it. Every extra line of the comp plan is a loophole being stress-tested by someone who needs the money and has nothing but time between calls. The test is simple and brutal: a rep should be able to look at a deal and know — in their head, in thirty seconds, without a spreadsheet, without emailing the comp administrator — exactly what the fucking deal pays. The moment they can't, one of two things happens. Either they stop optimizing entirely (you've destroyed the incentive you built the whole plan to create, nice work) or they bring in a "comp whisperer" — usually the most dangerous person on any sales floor, quiet, spreadsheet-literate, operating in the shadows — to reverse-engineer the plan and find the loophole that prints money while technically complying with every line.

2–3 components, maximum. Base + variable. Maybe variable plus one modifier. That's the plan. Anything more is job security for your comp administrator and confusion tax for everyone else.

HOW IT GOES TO HELL

The Guru — Chip Brennan — sold our VP of Sales a "neuroscience-backed multiplier matrix" at a revenue conference in Nashville for twelve thousand dollars plus a Masterclass. It has nine components, a quarterly momentum modifier, a "pipeline health coefficient," and a tier system that produces different outputs depending on whether Mercury is in retrograde. It is a stunning piece of intellectual architecture that absolutely no working salesperson can decode. Within sixty days the floor has bifurcated into people who've stopped trying to understand their check and one terrifying sociopath named Marcus who reverse-engineered the whole matrix in a spreadsheet and is now booking 0.5× services deals through a loophole in cell J12 that retires at 1.8× due to a rounding error in the multi-year tier calculation. Marcus is making $380K. The company is losing margin on every deal Marcus closes. Chip is on a podcast about "aligning incentive architecture with growth trajectories" and Marcus has already accepted an offer elsewhere, which he will activate the day the comp plan is corrected. Chip did not include this scenario in the twelve-thousand-dollar deck.

The Brilliant Jerk read the plan more carefully than you did — he does this every year, it is the first thing he does in January, he treats it like a puzzle and he is very good at puzzles. He found that the accelerator runs from 100% with no ceiling, and that quota retirement doesn't haircut deep discounts, so he's been offering 40% off list to accelerate close, printing 2× commission on each deal, and getting obscenely rich on transactions that lose the company margin. He is doing exactly what the plan says to do. That is the horror of it. The plan is the criminal. He is the weapon. You made him this way. Redesign the plan or keep writing the checks — those are your two options and they are both expensive.

The VP of Vibes changed the comp plan mid-year. There was a reason — there is always a reason, usually articulated with the phrase "we need to drive urgency" — but it doesn't matter. Reps had built their entire quarter's pipeline strategy around the old rules. Two of them were thirty days from closing deals that the old plan paid beautifully and the new plan barely rewards. Three top closers, mid-deal, quietly open their laptops at 10 p.m. on a Tuesday and start interviewing with competitors, because in sales comp there is one cardinal unforgivable sin, worse than a low OTE and worse than a sandbagged quota and worse than missing the year: moving the goalposts after the whistle. Trust, once broken by a mid-period plan change, does not heal with a steak dinner and a Q3 SPIFF on a product the reps don't believe in. It just doesn't. People remember what you took from them.

FIELD RULES

You get the behavior you pay for. Full stop, no asterisk, no exceptions. Reverse-engineer the plan from the behavior you want — start with the behavior, work backward to the incentive, then write the document. Never start with the document.

Simple beats clever every single time. If the plan doesn't fit on a napkin, it will be gamed by Thursday. Three components, maximum. The elegance you think you're building with nine components is the loophole you're actually building.

Fund the accelerator. Paying your top 20% obscene money for overperformance is the best return in the building. Do not cap it out of anxiety that someone might earn more than the CFO. The CFO should be thrilled. The overperformer just covered three misses.

Quota retirement is the steering wheel. Pay more for the deals you actually want; pay less, or nothing, for the junk. If reps are selling the wrong mix, look at the retirement table before you look at the rep.

Never, under any circumstances, change the plan mid-period. Lock it before the period starts, communicate it clearly, honor every line of it even if the business changes. The one unforgivable sin in comp is retroactive goalpost-moving, and it costs you your best people because your best people have options and a very long memory.

My attorney advises: every plan gets a signed written acknowledgment from every rep before the period starts, and a clause that explicitly reserves the company's right to modify plans at period boundaries. The unsigned plan is the one that ends up in arbitration at the worst possible moment, with a plaintiff's attorney asking you to explain verbally why the document the rep signed doesn't match the document in the CRM.

From the field: Comp is culture with a dollar sign. You can write "customer-first" on every wall in the building in forty-eight-point Helvetica. You can talk about long-term relationships and partnership and mutual success at every all-hands until you're hoarse and the slide deck is worn thin. But if the plan pays on a signed PO and stays completely silent on whether the customer ever receives value, ever renews, ever refers another buyer — if the plan makes zero distinction between a deal that lasts three years and a deal that cancels in ninety days — then your reps will sell to anyone who can fog a mirror and collect on the way out the door, and they will be doing the perfectly rational, mathematically correct thing your plan told them to do. The comp plan is the truth. The values poster is the aspiration. At 11:58 p.m. on the last day of the quarter, a rep with a number to hit and a mortgage payment clearing in four days will do the exact thing the plan rewards — precisely that thing, nothing less, nothing more — and they should. That's what you built. Build the right thing.