The customer success req sitting in your ATS was probably written in 2021. It asks for empathy, organization and a customer-first mindset. All true, all still necessary, and none of it the reason the role exists now.
Somewhere in the last three years, customer success stopped being the team that keeps customers happy and became the team that keeps revenue. The benchmark data says that shift has already happened. Most job descriptions have not caught up, and neither have most interview loops.
You are not hiring someone to protect the relationship any more. You are hiring someone to protect the number — and most of the market has never been trained to do it.
Retention is the growth plan now
SaaS Capital’s 2026 survey of more than 1,000 private B2B SaaS companies puts median net revenue retention at 103% and median gross revenue retention at 91% for companies between $3M and $20M in ARR. The 90th percentile reaches 117.9% NRR.
Read that pair together, because the pair is the whole story. The median company loses 9% of its recurring revenue outright every year and expands its way back to three points of net growth. Everything above flat comes out of the base you already own — and the person who owns the base is the CSM.
Segment changes the picture more than the headline does. Across ACV bands, companies with ACVs of $25,000 to $50,000 post a median NRR of 102%, with the top quartile at 111% and the bottom quartile at 97%.
| Retention benchmark, private B2B SaaS | Figure |
|---|---|
| Median net revenue retention | 103% |
| Median gross revenue retention | 91% |
| 90th percentile NRR | 117.9% |
| Median NRR, $25,000–$50,000 ACV | 102% |
| Top quartile, $25,000–$50,000 ACV | 111% |
| Bottom quartile, $25,000–$50,000 ACV | 97% |
A 14-point spread inside a single price band is not a market condition — those companies are selling comparable products to comparable buyers. It is an execution gap, and execution gaps are hiring decisions.
Eight in ten CSMs own revenue. Six in ten feel ready for it.
ChurnZero’s 2026 CSM Confidential Report asked CSMs directly whether they own revenue in some form — a quota, a target, or an informal shared responsibility. Eight out of ten said yes. Only six out of ten said they feel prepared to own it, a 23-point gap.
The reasons are specific rather than vague. 44% say sales and negotiation skills are what is holding them back, and nearly half report no commercial training of any kind in the last twelve months. Asked what would make them readier, the top answers split four ways: more time and less admin (39%), better selling and negotiating skills (39%), clearer expansion playbooks (38%), and clarity on rules of engagement with the sales team (31%).
For a hiring manager, that is the most useful fact on this page. The pool you are about to source from carries revenue responsibility on paper with no commercial training behind it. Which means commercial instinct is rare enough to be worth screening for deliberately — and trainable enough that you can hire for aptitude and build the rest.
The same research shows what closing the gap is worth:
| Received commercial training | No commercial training | |
|---|---|---|
| Feel prepared to own revenue | 72% | 50% |
| Happy in role | 70% | 58% |
| Agree CSMs can sell and advise | 91% | 63% |
Fifteen minutes of role-play inside a weekly 1:1 moves those numbers. The requirement is consistency, not a six-figure L&D budget.
One title, three genuinely different jobs
Most customer success reqs describe a blend of three roles and then interview for none of them. Decide which one you are hiring before the job description goes live, because the candidate profiles barely overlap.
| Adoption CSM | Commercial CSM | Technical CSM | |
|---|---|---|---|
| Owns | Onboarding, time-to-value, product adoption | Renewals, expansion, NRR on the book | Integrations, data health, technical escalation |
| Measured on | Time-to-first-value, adoption depth, gross retention | Net revenue retention, renewal rate, expansion pipeline | On-time implementation, technical churn, escalation volume |
| Usually comes from | Support, onboarding, implementation, enablement | AE, account management, or a sales-trained CS team | Solutions engineering, implementation, technical support |
| Hire when | Churn happens in the first 90 days | Churn happens at the renewal conversation | Churn traces back to a broken integration |
The mistake we see most often on customer success searches is hiring an adoption CSM and then handing them an expansion target two quarters later. That is not a performance problem when it fails. It is a scoping problem, and it costs you the person as well as the number.
If the job is genuinely commercial — a real quota, real negotiation, real pricing conversations — you are closer to an account executive profile than a support profile, and you should be sourcing accordingly. If it is genuinely technical, the bench you want overlaps with technical recruiting and solutions engineering, not with the CS talent pool at large.
Size the book before you write the req
The one-CSM-per-$1M-ARR rule is still quoted in planning meetings, and ChurnZero’s own team calls it a myth worth retiring. The dollar figure is not the problem. The problem is that the rule ignores your customer journey and assumes every customer costs the same to serve.
Build it bottom-up instead, before the req opens:
- Total the time each recurring touchpoint takes — success plan reviews, business reviews, product updates — times its frequency.
- Add one-time touchpoints like kickoff, smoothed across the base.
- Add monthly inbound volume, then subtract internal time. A useful rule of thumb is two-thirds of a CSM’s time for customer work, one-third for company work.
- Divide available hours by adjusted workload. Do it once per segment, not once for the whole base.
That exercise usually reveals the real reason you are hiring. Sometimes it is a headcount gap. Often it is that four people are carrying books nobody sized, and the fifth hire will land in the same trap. Getting it right before you open the req is also the cheapest way to avoid a slow, restarted search — and hiring delays in SaaS compound faster than most teams model.
AI took the admin, not the judgment
OnRamp’s 2026 survey of 150 CS and revenue leaders found 88% saying AI lets onboarding scale across customer tiers without adding headcount. That is the number that gets quoted in board decks.
The rest of the survey is the part that matters for hiring. Only 17% rate their team’s AI maturity as advanced. 95% describe their AI as mostly reactive — it explains what already happened rather than flagging what is about to. Only 30% use AI to detect a stalled onboarding proactively, and only 39% say they consistently hit their onboarding goals.
So the tooling has absorbed the reporting, the data-gathering and the administrative drag. It has not absorbed the judgment call in a QBR where a champion goes quiet, or the negotiation where a customer wants a 20% discount to renew flat. Hire for the second thing. Screen for AI fluency the way you would screen for CRM hygiene — as table stakes, not as the differentiator — and spend the interview on the commercial reasoning instead.
Teams building the systems layer underneath all of this are increasingly hiring for it explicitly, which is why GTM engineering and CS have started sharing a border.
How to interview for a revenue-owning CSM
Five questions, and what a real answer sounds like:
- “Walk me through a renewal you saved.” You want the mechanics — when they knew it was at risk, what signal told them, who they escalated to, what they traded. Vague relationship language means they were not close to the number.
- “Tell me about an expansion you sourced yourself.” Not one routed to them by sales. Listen for how they spotted it and how they handled the handoff, which is where most CS-to-sales friction actually lives.
- “Describe a churn you did not see coming.” The strongest candidates answer this fastest and in the most detail. Anyone who cannot name one has either not owned a book or is not being straight with you.
- “How was your book sized, and was it right?” This tests whether they think in capacity and segments or just worked the queue.
- “A customer asks for 20% off to renew flat. What do you do?” A pure commercial reflex test. There is no single right answer; there is a large difference between a candidate who negotiates on value and one who forwards the email to their manager.
Run those in a 45-minute loop and you will separate the CSMs who have carried a number from the ones who have carried a calendar. The distinction barely existed five years ago. In 2026, with the median company only three points above flat, it is most of the job — and it is why we scope every customer success search against the specific number the hire is expected to move, the same way we do across sales and sales leadership. The 2026 GTM hiring benchmarks are worth reading next to this one, because a slow process on a role this contested costs you the same way a slow AE search does.
