Stop Turning Your CSMs into Sales Reps in Sheep's Clothing: The CS Comp Playbook

The mistake most early-stage founders make is simple: they try to incentivize their way out of a structural problem. They think a higher commission will turn a CSM into a closer.

It won't.

It just turns them into a sales rep in sheep's clothing, and your customers can smell the commission breath from a mile away.

I've sat on both sides of that table, building CS orgs from scratch and inheriting comp plans that quietly broke the team's incentives. CSM compensation plans are harder to get right than sales comp. 

Sales is clean: you booked it, you get paid. CS is messier. You're paying people to influence outcomes (retention, adoption, expansion) that they don't fully control. Get the design wrong and you'll either underpay your best CSMs into burnout or hand them a quota that erodes the trust your product depends on.

Here's the playbook we're using with our clients to build and restructure their CS teams.

The 70/30 Split: Mechanics, Not Magic

For early-stage teams, I advocate for a 70/30 compensation split (Base/Variable), paid out quarterly.

Why 70/30? It's the "Goldilocks" zone. It provides enough stability for the CSM to remain a Health Guardian for the customer, but enough skin in the game to keep them focused on the business outcomes that drive revenue. The moment your variable comp climbs north of 35%, you start attracting candidates (and rewarding behaviors) that look more like sales. And that's not what your customers signed up for when they bought your product.

Pay it quarterly, not annually. Annual payouts are too far away to influence behavior, and they bunch up risk for both sides. Monthly is too noisy and creates panic over short-term events you can't control (like a customer's procurement cycle slipping a month). Monthly is also a massive headache for your RevOps team to calculate. They'll thank you for the quarterly payouts instead. Quarterly is long enough to smooth out lumpy renewals and short enough to feel like real reinforcement.

Build accelerators into the variable. A flat payout caps the upside for your top performers, and your top performers will notice. I typically design plans where hitting 100% of target unlocks an accelerator, say, 1.25x to 1.5x payout on every dollar (or every deal) above target. It costs you nothing if the team doesn't hit. And when they do hit, you're paying out of upside you wouldn't have had otherwise.

The two metrics that matter most for early-stage CS teams: GRR and LRR

This is where I see most founders get it wrong. They look at the public SaaS playbook, see "NRR" plastered everywhere, and decide their CSMs should be comped on NRR.

For an early-stage team, NRR is the wrong metric to put in your CSM comp plan. Here's why: NRR blends retention with expansion, and at your stage, you don't have enough deal volume for the expansion side of the equation to behave like a stable signal. One large expansion deal can mask a retention problem for two quarters. One churned logo can erase a great expansion quarter. You end up paying CSMs based on noise.

Instead, split the variable across two metrics: GRR (Gross Revenue Retention) and LRR (Logo Retention Rate).

GRR measures how much of last year's revenue you kept, excluding expansion. It's the cleanest read on whether your customers are renewing and at what dollar value. This is what your CSMs can directly influence through onboarding, adoption, and renewal motion.

LRR is the unsung hero of early-stage comp plans. Logo retention rate measures whether you're keeping your customers, full stop, independent of dollars. And here's why it matters more than founders realize:

If your customers are staying, even through layoffs, downturns, and budget cuts, it means your product is a vital part of their tech stack. That's the most honest signal of product market fit you'll ever get.

Your CSMs can't control whether a customer goes through a 30% RIF. They can't control whether a champion leaves for another company. But they can control how vital your product feels inside the account. They can control whether the customer's exec team understands the ROI when budget season hits. They can control the messaging, the QBR narrative, the executive alignment.

LRR is the metric that rewards CSMs for making your product indispensable. That's exactly the behavior you want to be paying for at SMB and early-stage scale.

A good starting split: 50% of variable on GRR, 50% on LRR. As you mature and your dollar retention becomes more meaningful, you can shift the balance toward GRR or layer in NRR.

Don't put expansion quota on your CSMs (yet)

I know. Every CRO and board member is going to push back on this. "If CSMs own the relationship, shouldn't they own expansion?"

Here's my honest take: the moment you make expansion a primary quota for your CSMs, you've turned them into AEs with worse comp. The CSM-as-trusted-advisor relationship is one of the most valuable assets in your business. It's why customers tell you the truth about why they're considering churning. It's why champions tell you what your product is missing. The minute the customer suspects your CSM is comped on selling them more, that channel of trust starts to close.

For early-stage teams, keep expansion ownership with your AEs, or hire a dedicated AM function once volume justifies it. Your CSM should be the one identifying expansion opportunities, building the business case, and making sure the customer is set up to expand successfully. The AE/AM should be the one running the commercial conversation.

Two ways to track CSM contribution to expansion (without making them sales reps)

You still want CSMs to lean into expansion. They're the ones with the relationship and the product knowledge. The trick is to reward the contribution, not the quota.

Here are the two structures I use:

Option 1: Closed Won opportunity count. Set a baseline target, say, 3 expansion opportunities sourced and Closed Won per quarter. Below the baseline, no payout. At baseline, full payout. Above baseline, a flat dollar accelerator per additional Closed Won opp ($X for every CW deal beyond the third).

This works well when your expansion deal sizes are relatively consistent. It's simple to understand, easy to administer, and rewards the behavior you want: nurturing expansions to close.

Option 2: Dollar-tiered payouts. Pay based on the size of the Closed Won expansion. Tier the payouts so larger deals pay materially more:

Use this structure when your expansion deals have wide dollar ranges. It aligns CSMs around prioritizing the bigger deals and nurturing them more aggressively, without putting a quota on their head.

The key distinction in both structures: this comp is for contribution to a Closed Won deal that an AE or AM commercially owns. The CSM isn't the deal owner. They're the trusted advisor who teed it up, kept the customer aligned, and made sure the use case landed. That separation matters.

When to split CSM and AM into separate roles

Most early-stage teams start with "full-stack CSMs," one person doing onboarding, adoption, renewals, and expansion. That's fine until it isn't. Here are the signals it's time to split:

The first signal is consistent expansion volume. If you have enough expansion motion that a dedicated AM could carry a $1M+ quota, hire the AM. Until then, AEs can hold the expansion bag with CSM support.

The second is renewal complexity. If your renewals are clean and software-like, your CSM can run them. The minute renewals get procurement-heavy, multi-stakeholder, and price-negotiation-driven, you want commercial DNA in the room. That's an AM, not a CSM.

The third is your CSMs are becoming the bottleneck for everything. When the CSM is doing the QBR, running the renewal, sourcing the expansion, and doing the technical enablement, your retention will quietly suffer because none of those motions are getting full attention.

Splitting roles isn't an org chart vanity exercise. It's how you protect the trusted-advisor relationship that makes the rest of the model work.

The TL;DR

If you're a founder building a CS comp plan from scratch:

The biggest mistake I see founders make is copying a comp plan from a Series C company when they're still pre-Series A. The right plan at $5M ARR looks different from the right plan at $50M ARR. Build the plan for the company you are, not the company you read about on LinkedIn.

And test your plan against this question before you ship it: if a CSM optimizes purely for their comp, does the resulting behavior make my customers more successful?

If the answer is yes, you've designed a good plan. If you have to squint, redesign it.