← BUILD LOGRETENTION + NRR
BUILD LOG · RETENTION, RUN PROACTIVELY

Churn was bigger than new business. So I built the system that saw it coming.

The year I took over, churn and downgrades were larger than everything the new-business engine brought in. We were filling a bucket with a hole in the bottom and congratulating ourselves on how fast we poured.

Steal this build ↓See the stackThe people part →
-28%
churn and downgrades, the year I took over
7 figures
churned-customer win-back the same read produced
66%
of renewals were getting zero renewal activity
69 to 180 days
first renewal touch, moved from late to early
THE STORY, FROM THE SEAT

Here is one I actually shipped. What happened, what I did, and the stack if you want to run it yourself.

The year started with a piece of math nobody wanted on the table: churn and downgrades were larger than everything the new-business engine brought in. We were filling a bucket with a hole in the bottom and congratulating ourselves on how fast we poured.

Renewals were handled reactively. A CSM found out an account was leaving roughly when the account told them. By then the conversation is not a save, it is an exit interview. And the audit made the scale of it undeniable: 66% of renewals were getting zero renewal activity. Not late activity. None.

You cannot save a renewal in the last thirty days. You can save it ninety days out, when the signals first turn. The whole job was to move retention from reactive to proactive, which means one thing in practice: see the risk early enough to act on it. So the work started with autopsies, not dashboards, because you cannot build an early-warning system until you know what the early warnings are.

The Sales Operator
Keep Building,
Heath
FOUNDER, THE SALES OPERATOR
THE BUILD IN FIVE MOVES · THE GIANT METHOD
  1. GChurn and downgrades outweighed everything new business brought in, and a CSM learned an account was leaving when the account said so.
  2. IGong autopsies found what predicted a loss, Amplitude flagged accounts turning cold, Salesforce held the book view, Deepline ran the renewal play.
  3. AAI runs the autopsies and the early-warning score. The CSM owns the save, months ahead of the date.
  4. NThe score drove a standing renewal review at 180 days, with a VALUE business case built from usage for every at-risk account.
  5. TChurn and downgrades fell 28%, and the same read fed a seven-figure churned-customer win-back.
RUN THE PLAY YOURSELFThis exact build is a playbook. Run it on your own data.
→
01 / HOW WE APPROACHED THE PROBLEM

Ran the loop.

The year started with churn and downgrades larger than the entire new-business engine, and renewals handled reactively: a CSM found out an account was leaving roughly when the account told them. You cannot save a renewal in the last thirty days; you can save it ninety days out, when the signals first turn. So the work began with autopsies, because you cannot build an early-warning system until you know what the early warnings are.

  1. 1
    Build the autopsies first

    I ran churn, closed-won, and closed-lost analyses to learn what actually predicted a loss. Not opinions, patterns. You cannot build an early-warning system until you know what the early warnings are.

  2. 2
    Give every CSM a book-of-business view

    One place that showed every account, its health, its renewal date, its risk. Not a spreadsheet someone updated when they remembered.

  3. 3
    Score the risk

    Account-health churn-risk scoring that flagged the accounts turning cold while there was still time to do something. The signal, not the surprise.

  4. 4
    Make renewal a motion, not a date

    A renewal-proactivity playbook so the work started months ahead, on the accounts the score flagged, in priority order.

02 / WHAT IT TOOK CROSS-FUNCTIONALLY

Nobody ships this alone. Here is who had to move.

Retention read like a CS problem and was not one. The early-warning system only worked because the signals, the money, and the second-chance motion each had an owner outside the renewal call.

PRODUCT
The usage signals the health score read.

An early-warning score is only as honest as its instrumentation. The behavioral data that flagged an account turning cold came from the product side of the house.

FINANCE
The renewal forecast and the discount line.

Every save conversation ran inside a policy. With the forecast and the discount boundary set, a save defended value instead of defaulting to a price cut.

SALES
The second-chance pipeline.

The same read pointed backward produced a seven-figure churned-customer win-back, and that revenue needed a pipeline and a motion like any other channel.

03 / HOW WE TURNED IT INTO A SALES MOTION · THE PEOPLE PART

The score saw the churn coming. The renewal review is how the CSM did something about it.

An early-warning score is only worth the conversation it triggers. The leadership half was turning the renewal from a 30-day scramble into a 180-day motion, and giving every CSM a business case to defend the number instead of a discount to protect it.

THE FRAMEWORK BEHIND IT
VALUE Business Case, run as a renewal review

Validate what is true today, articulate what good looks like, link real product usage to the outcome the customer actually bought, name the unique value they would lose, and put an execution date on it. The renewal becomes a value conversation, not a price negotiation.

  1. 01
    Started the renewal at 180 days, off the score

    The canary list drove a standing renewal review months before the date. The first touch landed when there was still time to fix the account, not when the clock had run out.

  2. 02
    Built the case from usage, not goodwill

    Every at-risk account got a VALUE business case tying the product usage the score could see to the outcome the customer signed up for. The number defended itself.

  3. 03
    Coached the save, then rehearsed it

    CSMs practiced the renewal and the objection they would actually get, price, low usage, a champion who left, before the live call, not during it.

Build the renewal case →
04 / WHAT I LEARNED

Not proof. Just what the build taught me.

  1. 01
    You cannot save a renewal in thirty days

    The save works ninety days out, when the signals first turn. By the last month the conversation is an exit interview, not a save.

  2. 02
    Learn the warnings before you build the warning system

    Churn, closed-won, and closed-lost autopsies came first. Patterns, not opinions, told the system what to flag.

  3. 03
    A healthy-looking book is the trap

    90% of the renewal book read fine right up until it did not. The warnings sat in the data, unread.

  4. 04
    A score is worth the conversation it triggers

    The early-warning score mattered because it started a renewal review months before the date, with a business case to defend the number instead of a discount to protect it.

05 / THE WORKFLOW

The runnable version. Copy it into your stack.

GongSalesforceAmplitudeDeepline
[ SALES OPERATOR ]

Proactive Retention

Heath Barnett · Sales Operator
G / Groundthe problem that started all of it, with the receipt and the cost
Risk

The save had already become an exit interview.

A CSM found out an account was leaving about when the account told them. Too late to save it.

Signal

The book looked healthy on paper.

90% of the renewal book read fine right up until it did not. The warnings were in the data, unread.

A / Assignthe build, block by block, and who owns each one. Open a step to see it run.
HUMAN + AI, IN THE LOOP
AI runs the autopsies and the early-warning score; the CSM owns the save, 90 days ahead of the date.
Learn the warnings
Gong01
Run churn, won, and lost autopsies
Details
Patterns, not opinions. What actually predicted a loss from the calls.
AI
Amplitude02
Read the health signal early
Details
Flag accounts turning cold while there is still time to act. The signal, not the surprise.
AI
See the book
Salesforce03
Build the book-of-business view
Details
Every account, its health, renewal date, and risk in one place, not a spreadsheet updated from memory.
Human
Amplitude04
Score the churn risk
Details
Rank the accounts turning cold so the save starts 90 days out, not 30.
AI
Act 90 days out
Deepline05
Run the renewal play in priority order
Details
The work starts months ahead, on the accounts the score flagged, in priority order.
AI
The CSM06
Own the save conversation
Details
A save while there is still time, not an exit interview.
Human
N / Normalizethe motion that made it stick
A workflow without a motion is dead. This is how it became the way the team works.
1

The renewal starts at 180 days

The score drove a standing renewal review months before the date, so the first touch landed while there was still time to fix the account.

2

The case is built from usage, not goodwill

Every at-risk account got a VALUE business case tying the usage the score could see to the outcome the customer signed up for. The number defended itself.

3

The save was rehearsed, not improvised

CSMs practiced the renewal and the objection they would actually get, price, low usage, a champion who left, before the live call.

T / Tie backthe result it drove, and how they know
RESULT · 01

90-day SLA

every renewal gets a logged touch, auto-flagged if not.

RESULT · 02

Reactive to proactive

the save moved 90 days ahead.

RESULT · 03

Risk seen early

accounts flagged while there is still time.

OUTPUT · 01

Churn autopsy patterns

What actually predicts a loss.

OUTPUT · 02

Book-of-business view

Every account, health, renewal, risk.

OUTPUT · 03

A 90-day renewal play

Work that starts before the date.

The Sales Operator
06 / HOW YOU DO IT TOO

Here is what I built. Here is how you build it.

The whole thing installs as one plugin. Or grab the skills a la carte. Everything on this page, runnable, on what you paste today.

✓ EVAL PASS · 4/4
THE PLAYBOOK · ONE INSTALL
Save an at-risk renewal

4 skills chained into one runnable play. Installs as a single plugin, no copy-pasting each skill. It runs on what you paste; connect your stack to go live.

SEE THE FULL PLAYBOOK →
SEE IT RUN IN CLAUDE
EXAMPLE CHATPriya, a CSM, running the loop in one sitting
P
I've got a renewal in 45 days and something feels off. Can you give me a straight read on whether it renews?
S
Step 1· Renewal Health Analyst
Combined adoption, champion stability, support sentiment, and competitive signals into one verdict. Adoption dropped a third since Q1, the original champion left in March, and two support tickets went unresolved. Verdict: at risk.
adoptiondown 34%
championgone
Receipt
Renewal verdict: at risk, driven by champion loss and a 34% adoption drop.
P
Walk me through what's actually been happening in the relationship. When did it start slipping?
S
Step 2· Conversation Analyst
Put every meeting, reply, and sequence touch on one timeline. The slip started right when the champion left in March, replies got slower, and the new stakeholder has never joined a call. That's the champion-drop moment.
Receipt
Timeline shows the drop-off began at the March champion exit, 3 unanswered touches since.
P
Okay, I need to actually save this. What's the play?
S
Step 3· Save Play
Diagnosed the real cause, not the stated one, this is an orphaned account with no internal owner, not a product gap. Matched levers to that: re-anchor with a new economic sponsor and rebuild value, not a discount. Built a multi-touch save with exit criteria.
Receipt
Save play built on the real root cause, 4-touch sequence to re-sponsor the account.
P
They're going to ask for a discount to stay. How do I handle the renewal terms?
S
Step 4· Renewal Negotiation
Wrote the value recap, laid out price and term options, and built a concession ladder so any discount buys us a multi-year term or an exec sponsor. Set the floor that protects net retention and scripted the answer to 'we want a discount.'
floorflat renewal
trade2-yr term
Receipt
Renewal plan protects net retention, save path closes the renewal flat on a 2-year term.
THE SKILLS IT USES
Get a renew-or-not verdict on every renewal before the meeting.
renewal-health-analyst
See your whole book in one rollup.
book-of-business-analyst
Read the whole relationship at a glance.
conversation-analyst
THE OTHER HALF · LEAD THE TEAM
Build the renewal case→
I write up one of these a week. Free, receipts only.
RELATED BUILDS
The Expansion Score
For every ten seats they paid for, about seventeen people were using it
Reporting That Drove Action
One weekly read that drove sales and CS, not a rear-view deck
THE MARKET ON THIS
The craft behind it: VALUE Business Case (PLAN Selling)Customer health scores: metrics, models and tools (Gainsight)Predict churn with a customer health scoring system (Union Square)