You can hire a CSM who explains exactly why the webhook broke, maps every failed sync, and files a beautiful escalation — and still loses the renewal. You can also hire one who trades a twelve-month extension for a discount nobody authorised. Both failures look identical on a resume. Neither shows up in a conversation about “customer obsession.” The only way to tell them apart before you sign an offer is to hand the candidate a churn-risk account — usage curve, ticket history, contract terms — and ask for the save plan in writing.
Ninety minutes. One account. A published rubric. That’s the whole test, and it does more work than three rounds of behavioural questions.
Why the resume can’t tell you this
Commercial instinct in customer success is learned behaviour, not a default setting. Only 44% of CS teams are compensated for expansion — down from 57% the year before — which means more than half the candidates in your pipeline have spent their careers in orgs where pricing, trading and closing were structurally somebody else’s job. They’re not hiding commercial skill. They were never paid to build it.
They’ll tell you so themselves. Asked to name the skills gap standing between them and success, 42% of CSMs cited sales and negotiation — the single most pronounced gap in the study, ahead of business skills and analytical skills. And 59% say their company values sales more than customer success, against just 6% who say the opposite. That’s the environment your candidates are coming from. Commercial reasoning has to be elicited in an exercise, not inferred from a job title that happened to include “renewal ownership.”
The resume is worse than silent — it’s gamed. Among job seekers trying to get past AI screening, 40% admitted adding skills or keywords they didn’t have. “Owned $4M renewal book, 97% GRR” costs nothing to type. A save plan costs ninety minutes and cannot be faked by someone who has never written one.
If you don’t test commercial thinking, you’re not assessing it — you’re hoping for it.
What the take-home actually contains
Give the candidate a single account packet, no more than four pages:
- Contract facts. ACV, term, renewal date 70 days out, seat count, discount history, who signed, auto-renewal clause, notice period.
- Usage data. Monthly active seats against licensed seats over 12 months, feature adoption by module, a clear inflection where usage dropped after an org change.
- Ticket history. Twelve to twenty tickets, including two severity-one incidents, one unresolved integration complaint, and a feature request that was declined.
- Relationship map. The champion who left, the new VP who has a competitor relationship, the economic buyer nobody has met.
- One sentence of ambiguity. “The CFO has asked all software owners to justify renewals above $100K this quarter.”
The ask: a save plan of no more than two pages — diagnosis, intervention sequence with owners and dates, the commercial recommendation, and the one thing they’d escalate to you in week one.
Time-box it honestly and say so in the brief. Candidates are not allergic to work samples; 96% of job seekers said they wanted a way to show employers something more clearly than their job application allows. A scoped, paid-respect, ninety-minute exercise with a visible rubric is a draw. An open-ended “build us a QBR deck” with no time limit is the drop-off risk, and deservedly so.
The three answers that separate technical from commercial
Every save plan answers the same three questions. How a candidate answers them is the entire read.
1. “What is actually wrong here?” The technical CSM diagnoses the product: the integration is failing, adoption of module three never happened, tickets cluster around a known defect. All true. The commercial CSM diagnoses the budget: the champion who justified the spend is gone, nobody has re-sold the value to the new VP, and the CFO’s memo means this renewal is now a defence of business case, not a defence of uptime. Product problems are often the symptom; the missing internal sponsor is the disease.
2. “What are you willing to trade?” This is the one most candidates skip entirely. A technical answer offers free services: more training, a dedicated solutions engineer, an executive escalation. All cost, no capture. A commercial answer names a trade — multi-year term for the discount they’re about to ask for, a seat reduction in exchange for a two-year commitment, a reference or case study against a services credit, a price hold that expires if they don’t sign by the renewal date. Watch whether they know what they have to give and what they expect back.
3. “What’s the walk-away?” The technical CSM will fight for every logo. The commercial CSM tells you the floor — below this discount, the account is dilutive and we let it go — and recommends a downgrade path instead of a death spiral. They also tell you what the renewal is worth in the save plan: 40 seats at a 20% discount beats 90 seats churning, and they say the number.
| Question | Technical CSM answer | Commercial CSM answer |
|---|---|---|
| What’s wrong? | Integration failures, low module adoption, open sev-1s | No internal sponsor; value case never re-sold to the new VP |
| What do you do first? | Escalate to engineering, book a training series | Get a meeting with the economic buyer inside 10 days |
| What do you trade? | Free services, an exec apology, more enablement | Term length, seat count, reference rights, a dated price hold |
| What’s success? | Tickets closed, usage recovered | Renewal signed at a defined floor, with a downgrade path named |
| What do you escalate? | The defect | The discount authority and the CFO’s deadline |
Neither column is wrong. A pure commercial answer with no product diagnosis is a renewal rep, and you’ll hear it in the vagueness about why usage fell. What you want is both, in that order, in one document. Score for the combination, not the flavour.
The rubric — publish it with the brief
Send the rubric in the same email as the packet. Candidates calibrate to it, you stop reading impressionistically, and your hiring panel argues about evidence instead of vibes.
| Dimension | Weight | What a 5 looks like |
|---|---|---|
| Data reading | 20% | Names the inflection point in the usage curve and ties it to the org change, not just “adoption is low” |
| Root-cause diagnosis | 20% | Separates product symptom from sponsorship cause; says which one kills the renewal |
| Commercial recommendation | 25% | Names a trade, a floor, and a dated offer — with the arithmetic shown |
| Sequencing and ownership | 15% | Week-by-week plan to the renewal date, with named owners and internal dependencies |
| Stakeholder strategy | 10% | Has a specific plan to reach the economic buyer, not just the user base |
| Written clarity | 10% | Two pages a CRO could act on without a meeting |
Use the same rubric for every candidate on the req, and have two reviewers score independently before they talk. If the panel disagrees by more than one point on commercial recommendation, that’s the discussion worth having — and the thing that most often predicts whether a CSM defends gross retention or just documents its decline.
Size the account to the segment you’re hiring for
A save plan for a $9K SMB account and a save plan for a $400K enterprise account are different jobs, and a candidate who’s strong at one may be useless at the other. An enterprise CSM carries a $2.5M book, more than double the $1.2M book carried by an SMB CSM — which means the enterprise exercise should test depth on one account and the SMB exercise should test triage across a segment.
The stakes also differ. Companies with ACVs above $250,000 show median gross revenue retention of 95%, against roughly 91% below $250,000. Gross retention is the number a save plan actually moves — and the higher-ACV book is both the better-performing one and the one where a single blown renewal is visible in the board deck. Make the take-home account look like the accounts they’ll inherit in week one. If your mid-market team is the growth bet, build the packet from a real mid-market churn case with the names scrubbed.
For what those books pay and how comp is structured across segments, our salary benchmarks and the broader hiring guides are the starting point before you write the offer.
Treat the take-home as a signal, not a verdict
Work samples are good. They’re not oracles. The 2021 re-analysis of personnel selection research cut the work-sample validity estimate from .54 to .33 — a meaningful downgrade on a generation of hiring folklore. So don’t let the save plan carry the whole decision.
What carries it is what you do next. In the same re-analysis, structured interviews scored .42 against .19 for unstructured ones. The highest-value forty minutes in your whole process is a structured debrief built on the rubric: walk the candidate through their own plan, ask why they chose that trade, push on the floor, change one fact — “the CFO’s deadline just moved up three weeks” — and watch them re-sequence live. That’s where you find out whether the commercial reasoning is theirs or borrowed from a template.
Run it the same way for every candidate. Same questions, same order, same scorers. The rubric is what makes the structure possible.
How this fits a real CSM process
Four steps, nothing wasted:
- Screen on book size, segment and whether they’ve carried a renewal or expansion number — ask for the quota, the attainment, and who signed the order form.
- Hiring manager conversation on one real account they saved and one they lost, with numbers.
- Take-home — ninety minutes, packet and rubric sent together, 72 hours to return it.
- Structured debrief — rubric-scored, two reviewers, scores logged before discussion.
That’s it. No panel of seven, no mock QBR to a room of strangers. Because the market moves in days, not weeks, and a process that takes a month loses the candidate who can actually do the job to a company that decided faster.
If you’re staffing the renewal motion end to end, the same logic applies next door: test account executives on a deal they have to price, and sales leaders on a forecast they have to defend. The principle is constant — make the candidate produce the artefact the job actually produces. For CSM searches specifically, that’s exactly how we screen top-tier talent before you ever spend an hour on an interview.
