Brief 01 · Renewals · Nov 2026
Green accounts that still churned
A plain breakdown of one customer success postmortem: what it claims, what is only the author's word, and a check a small team can run.
Source"What my CS team was missing" by Manuel Faust (founder of a company that sells an early warning product, and the article ends with a sign up pitch for it.), The Next Web (contributed article, not TNW reporting), 16 February 2026. Read the original. Full credit at the bottom.
Read the captions
- Green everywhere. Gone in 90 days.
- Both green. NPS 72.
- 23 tickets. Closed one by one.
- 3 of 5 power users changed profiles.
- Champion title, one level down.
- A health score shows where an account was.
- Tickets, last 90 days.
- Who your champion is.
- What changed since the last call.
- 60 days out, not 6.
- A few dozen accounts, a Friday afternoon.
- Renewal windows: 30, 60, 90 days.
- 7 day trial. No card. cshubb.com
enterprise accounts, within the same 90 days
Author's accountat one account over four months, each closed on its own
Author's accountchanged their profiles in a two week window
Author's accountat one of the two accounts, still green
Author's accountThe case
The case in one minute
Two enterprise accounts left the author's company within the same 90 days. The author calls it a $2M quarter, without saying whether that means contract value or revenue, or how it split between the two. Both accounts showed green in every internal system, and one had an NPS of 72.
The author says the postmortem found a coverage gap: the warning signs sat in places the team's process did not look. He does not describe it as an execution problem.
What was found
What the two postmortems found
Account one
Engineers filed 23 tickets about API latency across four months.
23 small tickets, each closed on its own. Read together, they showed rising frustration. None became critical.
Account two
Three of five power users updated their profiles in the same two week window.
One of them started posting about a competitor's product.
The argument
What a health score and an EBR can and cannot see
- An executive business review captures what the customer chooses to say in the room.
- A health score summarizes what the team measures, and it lags behind what is happening on the account.
- The buyer and the person who renews are often different people, so a happy champion tells you little about the budget holder.
- The signals live in five or six systems, and nobody's job is to read them together.
The framework
The three layer signal framework
The author groups the signals into three tiers. The article gives no numeric thresholds for any of them.
Support ticket patterns
A steady trickle of resolved tickets from one team can say more than a single critical escalation.
People changes
The person who bought and the person who renews are often not the same.
Competitive exposure
Shows interest in alternatives before it reaches a conversation with you.
A note on tier 3
Watching public professional activity is a sensitive practice. If your team does this, check local privacy law and your own policy first. We do not recommend it as a default and CSHUBB does not do it.
Key facts
The facts, with their status
| Item | What the article states | Status |
|---|---|---|
| Accounts lost | Two enterprise accounts within 90 days | Author's account |
| Value | "$2M quarter", unit not defined | Unverified |
| NPS | 72 at one account | Author's account |
| Tickets at account one | 23 over four months, all resolved individually | Author's account |
| Profile changes at account two | 3 of 5 power users in two weeks, champion title one level lower | Author's account |
| Accuracy claim | About twice a health score, 300 accounts, 18 months | Unverified, no method |
| Manual effort | About 90 minutes per account per week | Author's estimate |
| Scaling limit | A few dozen accounts by hand, automate at about 80 to 150 | Author's estimate |
| Author interest | Founder of a company that sells the fix, and the article ends with its sign up pitch | disclosed. |
The one number we do not repeat
The author says one rate of increase in tickets predicted churn about twice as well as a composite health score. The article does not say whose accounts, how churn or accuracy was measured, what the baseline was, or how many accounts churned. We cannot verify it, so this brief and its video do not use it.
Doing it by hand
The manual approach and its limit
The author describes one unnamed team whose managers logged signals from six sources every Friday, about 90 minutes per account per week. He says the method stops working past a few dozen accounts. The figures and thresholds are his own estimates, with no source.
Our reading: the defensible takeaway is that a hand check is realistic for a small book and unrealistic for a big one.
What it recommends
What the article recommends doing
- Keep the existing process: EBRs, health scores, stakeholder maps.
- Add a signal layer on top using the three tiers.
- Manual teams do a weekly Friday review. Automated teams get an alert when something changes.
- Match the response to the signal. A sponsor who stopped logging in needs a different move than a competitor's contact appearing in a sandbox.
What the article is also doing
It closes by promoting the author's own product, with a form for a one page executive brief of 10 accounts. Those product claims are not independently verified, and the article promises to show how both churned accounts would have looked in the product but does not. That does not make the case wrong. It means the claims should be tested, not trusted.
Our check
A simple pre renewal check
Built from the parts of the case that need no special tool. Our own, not from the article. Run it for every account renewing in the next 90 days.
- TicketsCount tickets in the last 90 days against the 90 days before. Note whether one team keeps filing.
- PeopleConfirm your champion and budget holder are still in role. Note what changed since your last call.
- NotesWrite the findings next to the renewal date, with a name and a next step.
- TimingDo this at 60 days out, not at 6.
- ScaleA few dozen accounts fits a Friday afternoon. Past that, you need one place for it all.
Where CSHUBB fits
What CSHUBB does for this
- Renewal windows at 30, 60 and 90 days.
- Configurable health scoring, separate from NPS.
- Playbooks triggered by health, renewal window, NPS or alerts, with tasks assigned automatically.
- Integrations with Zendesk, HubSpot, Jira, Gmail, Google Calendar and Microsoft 365, plus any webhook source.
- 7 day trial, everything unlocked, no card.
Credits
Source and credit
Source and credit
- Article
- "What my CS team was missing"
- Author
- Manuel Faust, founder of a company that sells an early warning product, and the article ends with a sign up pitch for it.
- Publisher
- The Next Web (contributed article, not TNW reporting)
- Published
- 16 February 2026
- Read the original
- thenextweb.com/news/what-my-cs-team-was-missing
- Last checked
- 15 November 2026
We use the case as a prompt for a useful check. We have no connection to the author. Every claim below is marked as stated in the article, the author's own claim, or our own check. The author's own site lists The Next Web among its customer stories.
We summarize and comment on this article under fair use and link to the source. Check the original for the author's full wording. The numbers are the author's and have not been audited by us. Last checked 15 November 2026.