The NPS Autopsy Playbook 2026
Ask when an account crosses a line against the benchmark, not on a calendar: users well past it get the advocacy question, accounts where a gap is opening get a friction question, new accounts that reached their first goal get a value check. Each answer is read against usage, and the pairing puts a move on the CRM record – or no move, with the reason noted.
Your happiest customers use the product the way your successful customers do, and say so when asked. The accounts worth a closer look are the ones where those two disagree – a heavy user who's unhappy, or a fan who barely logs in. A quarterly survey read on its own can't tell them apart: the score never meets the usage.
This play asks when an account crosses a line against the benchmark, and reads each answer against the account's usage. The pairing decides the move on the CRM record – or no move, with the reason noted.
Measure it on response rate, days from answer to the owner's decision, how many thin-use accounts reach the benchmark within eight weeks, and the renewal rate of the accounts you asked.
How it works14 steps
01SignalLet the benchmark pick the moment
Build the benchmark as in the benchmark play: customers in the same segment, 12+ months in, renewed at the same or higher revenue (aim for 20). Take the median of three or four usage measures – active users, weekly use of your core features. Your CS lead approves it.
Each week, flag accounts that cross one of three lines (defaults to tune):
- Well past: at or above the median on every measure for 30 days. Ask its two or three most active users.
- Gap opening: active users down a fifth in four weeks, or one more measure below 75% of the median. Ask the most active user of the feature that went quiet.
- Goal reached: a new account hits the first usage goal your successful customers reached in their first month. Ask whoever got it there.
Skip anyone surveyed in the last 90 days, and any account with an open save or renewal conversation. Run it as a weekly saved query in PostHog, Amplitude or Mixpanel, or a spreadsheet. Under about a hundred accounts, do it by hand.
02DecisionMatch the question to the moment
The line decides the question: one or two, in-product, after the user finishes a task. Your team chose to automate these asks, and each answer lands on the account record through a Zapier step or Sprig webhook to HubSpot.
03ActionAsk users well past the benchmark the advocacy question
The classic 0–10 "would you recommend us?", with a free-text follow-up. Only these users see it: you can act on their answer the same week.
04ActionAsk accounts with an opening gap what got harder
No 0–10 scale. A 1–5 "how well is this working right now?" and one free-text question about the feature their usage says they dropped.
05ActionAsk new accounts whether the first goal paid off
Within a week of the goal: yes, partly or no. A no this early is cheap to fix.
06ScoreRead each answer against usage and the records
Sort each answer: positive (9–10, 4–5, yes), neutral (7–8, 3, partly) or negative (0–6, 1–2, no). Pair it with usage: at or above (every measure at 75% of the median or better – default) or below. Then check 60 days of tickets and CRM notes, and the renewal date and ARR. A known outage changes what an answer means.
07DecisionDecide whether the answer adds a move
Leave it alone when the answer and usage agree: a positive or neutral answer from an account at or above the benchmark, unless it's a 9–10 on the advocacy question. Everything else picks a move.
08ActionLeave it alone, and write down why
Set Recommended move to No move, paste the verbatim, write the reason in one line, and set Review date 90 days out, when the user can be asked again. The leave-it-alone play covers the call.
09DecisionPick the move the pairing supports
- Frustrated: any negative answer. At or above the benchmark, it's an invested user telling you what to fix; below it, the owner may open the save play afterwards. Owner conversation.
- Thin use: positive or neutral answer, usage below the benchmark. They like you but aren't getting the value yet. Guided session.
- Advocate: 9–10 on the advocacy question, usage at or above. Reference ask.
10ActionDraft an owner conversation with the answer attached
A task for the owner with the verbatim, the usage behind it and the note below. Default: talk within five business days.
11ActionDraft a guided session to close the usage gap
Offer thirty minutes on the feature where they trail the benchmark, using their own work.
12ActionSend the reference ask your team chose to automate
Advocates get the ask from Customer.io within a day, logged on the HubSpot record. Start with drafts the owner sends; once they've been right for a few weeks, automate. If you run the advocacy play, send advocates there instead – one ask per person. This is a private reference ask. Don't forget to ask your customers to leave you a review on public sites too.
13Human stepYour team decides and runs the move
The CSM or AM runs the move as drafted, changes it, or passes. They own every customer conversation. Before acting, record the success measure and review date: a call held is activity, usage back at the benchmark is progress, a renewal or expansion is the business result.
14OutcomeReview what changed, then sharpen the benchmark
On the review date, record: did usage move, did the next answer change, and what else played a part? The verbatims answer that last question best. If nothing moved, change the move, stop it, or give it time (the eight-week review has the method).
Each quarter, rank the verbatims by feature, with ARR attached, for your product team, and check the pairings against renewals: if happy thin-use accounts renew anyway, the benchmark measures the wrong thing. If your board wants an NPS trend, keep a small random-sample survey for it – these answers come from chosen moments, so they don't make a fair average.
Setup and templates
Sources: usage from PostHog, Amplitude or Mixpanel; survey answers from Sprig (or a form you send by hand); tickets from your help desk; renewal date and ARR from the CRM. Account-level numbers need your analytics tool's account add-on (group analytics in PostHog, Accounts in Amplitude, Group Analytics in Mixpanel) and a group call in your app. Save the query, then subscribe to it (in Amplitude, add it to a dashboard and subscribe to that) so the result lands in your inbox or Slack on the schedule you set.
Sprig: trigger each Sprig survey on the task's completion event, sent from your app. Each week, send the chosen users to Sprig as a group or attribute, and filter the survey on it. Sprig answers reach Salesforce or Pipedrive through Zapier. With PostHog, a question plus a follow-up needs the paid Surveys plan; send answers to HubSpot through PostHog's HubSpot destination (contact properties; a HubSpot workflow copies them to the company).
Company properties in HubSpot (Starter or above; free HubSpot allows 10 custom properties in total), or your CRM's account object:
| Property | Type | Filled by |
|---|---|---|
| Survey moment | Dropdown select: Well past · Gap opening · Goal reached | Weekly check |
| Survey answer | Dropdown select: Positive · Neutral · Negative | Zapier or Sprig webhook, from the survey response |
| Verbatim | Multi-line text | Zapier or Sprig webhook, from the survey response |
| Usage vs benchmark | Dropdown select: At or above · Below | Weekly check |
| Recommended move | Dropdown select: Owner conversation · Guided session · Reference ask · No move | Weekly check |
| Move reason | Single-line text | Weekly check |
| Success measure | Single-line text | Account owner |
| Review date | Date picker | Weekly check, then account owner |
| Review result | Dropdown select: Usage and answer improved · Usage improved · No change · Stopped | Account owner |
Add Last surveyed (Date picker) on the contact, filled when Sprig sends the survey (Sprig's Study sent webhook), or by hand. It drives the 90-day skip; set Sprig's recontact waiting period to 90 days as well.
The automation here uses HubSpot workflows (any Professional hub). On Free or Starter, run the same saved view and create the tasks by hand. The weekly list is a saved view: Recommended move is known and Review date is this week or overdue. In HubSpot that's two filter groups: Review date is This week, or Review date is before Today. Each move creates a task for the owner, "Survey answer: [move]", with the verbatim and draft in the body. Slack can notify the owner; the record is where the work lives. HubSpot's Slack app sends task notifications; custom messages need a workflow (Professional or Enterprise). Customer.io writes each send to HubSpot with a webhook step that calls HubSpot's API. If the From address is a HubSpot user's, a BCC to your HubSpot logging address also works. Apply these checks to marketing emails sent by hand or automatically: permission to send, accurate sender details, required address information, an easy unsubscribe and shared suppression checks. Keep required service notices separate from marketing.
Advocacy question (Well past): "How likely are you to recommend [product] to a peer? 0–10" · "What's the main reason for your score?"
Friction question (Gap opening): "How well is [product] working for your team right now? 1–5" · "What's got harder about [feature]?"
Value check (Goal reached): "Did [milestone] get you what you needed? Yes · Partly · No" · "What were you hoping for?"
Owner conversation (Frustrated):
Hi [name] – thanks for being straight with us about [verbatim topic]. It matters to us that [feature] works for your team. Could we take twenty minutes this week to go through what's in the way?
Guided session (Thin use):
Hi [name] – glad [product] is working for you. Teams like yours get the most out of [feature]. Would thirty minutes help? We'll set it up with your own [work item].
Reference ask (Advocate):
Hi [name] – thanks for the [score]. Would you take one reference call with a team considering [product]? We'll check the timing with you first.
No move: "No move – [answer] matches usage; [reason]. Recheck [date]."
How Accoil fits
Accoil is a customer-success consultancy with our own tooling, and this is a play we run with you, not for you. We build the benchmark with your team, and you approve it. Every week we check every account in the agreed segment, and the recommended move, the answer and its evidence land on the account record in your CRM – we recommend HubSpot – at a pace your team can actually run. Your team makes every call. Nobody from Accoil contacts your customers. After the agreed window we review what changed together, and you keep the playbook, the benchmark and the analysis.
Running it yourself? Swap HubSpot for Salesforce or Pipedrive, Sprig for PostHog, and Customer.io for Ortto or Userlist. On Salesforce, the field writes need API access (Enterprise and up, or an add-on on Professional). On Pipedrive, automations and date triggers need the Growth plan. The play stays the same.
Thirty minutes with Kate, Simon or Peter on one example segment: what a successful customer looks like, where the gaps are, and whether there's work worth doing this quarter.
Talk to a founder →Keep reading
Second-Value Seeding Playbook 2026
Your best customers show you which feature comes after the first. This play checks every account against them each week, finds the feature customers like it added next and kept, and puts the move on the account in your CRM: an in-app prompt and a short story, the CSM on the next call – or no move, with the reason noted.
Community-Led Retention Loop Playbook 2026
Your advocates are already in your data: the accounts and people using the product well past what your successful customers do. This play takes them from the weekly check, looks for a reason to wait, and puts one ask on the account in your CRM: a community invite your team chose to automate, a CSM's personal ask – or no move, with the reason noted.
Monday-Morning Account Triage Playbook 2026
Your best customers show you what good looks like. Every Monday, this play compares each account in a segment with that benchmark, ranks what each gap is worth, caps the list at what your team can run, and puts the week's moves on the account records in your CRM: close a gap, grow the upside, or leave it alone with the reason noted.
Every playbook, one download
All 32 workflows as print-ready playbooks — diagrams included. Plus every new workflow as we publish it.