Build a Customer Renewal Presentation Around the Work That Actually Happened

The old sales deck is sitting in the shared folder, and it is tempting. It already has the promise on slide three, the logo, the architecture diagram, the quote from the customer's director of operations. Pasting it at the front of a renewal presentation makes the next year look like a continuation of the first pitch.

That is exactly the problem. The customer has spent the last twelve months inside the gap between that promise and what they actually got. They will spend the meeting quietly comparing the two, and you will spend it wondering why the room feels cold.

A renewal presentation should do two things, in this order. First, give an accurate account of a period that has already happened: what you agreed to do, what was delivered, what people used, what was observed, and what is still open. Second, propose something about the period that has not happened yet — which may be a continuation, a narrower scope, a different service, or a recommendation to stop part of it.

The second part must not rewrite the first. That is the whole discipline. If the engagement produced fewer results than the signed scope implied, the deck says so, and the proposal changes to match. Renewal is not evidence that the original outcome occurred.

1 — Return to the purpose and period being reviewed

Start by finding the goal as it was written at the time, not as you remember it. That means the signed scope, the proposal the customer accepted, the email where they said yes. Quote it, or paraphrase it faithfully, and mark where it came from. If the objective was phrased loosely back then, that looseness is a fact about the engagement, and it is worth saying out loud rather than tightening the language retroactively.

Then name the review window. Not "this year" — the actual dates, and whether they cover the full contract period or only part of it. If you are reviewing six months of an eighteen-month engagement, say six months. If a quarter of the support record is missing because of a ticketing migration, say that too. A deck that quietly presents an incomplete period as a complete one invites the customer to correct you, which is a bad use of the meeting.

What changed during the period is the third item, and this is the one you cannot answer alone. A customer sponsor left. A market shifted. Volume doubled. A system you built something on top of is being replaced. Put your understanding of those changes on one slide and ask the customer to correct it. "Here is what we understood the goal to be in March. Here is the period we're reviewing. Here is what we think changed on your side. Tell us where this is wrong."

One more thing belongs at this stage: the invitation. Say what the meeting is. A routine review that arrives with an unexpected renewal ask teaches the customer to treat every future invitation as a sales event, and you will not get that trust back cheaply. If the conversation is a strategic review, call it that. If it is a renewal negotiation, call it that. If it is both, say both, and say which part comes first.

2 — Distinguish delivered, adopted and useful

These are three separate statements and they need three separate rows, because they call for different responses next period. An enabled capability that nobody uses is a configuration or training problem. A capability that is used constantly with no measurable effect is a design problem, or a measurement problem. Only the third row — the Useful row — tells you whether the thing you sold actually did anything.

Keep the source and its limits attached to any claim that matters. "The register shows 1,180 worker records with current data" is a different kind of statement from "credential compliance improved." The first is a count from a system you control. The second is an interpretation that may need the customer's agreement.

There is a fourth case, and it is the most common one in renewal decks: the capability was used, and nobody ever agreed how you would know whether it worked. That is not a success with a thin evidence section. It is an unmeasured engagement, and the honest thing is to name it and propose how to measure it.

Here is the schematic, unpopulated, for a single slide. Build this before you build anything else:

1. Why we're here, and the period under review  _______
2. Agreed purpose, as written                    _______  (source: _______)
3. What we delivered                             _______  (source: _______)
4. What was actually used                        _______  (source: _______)
5. Outcome observed                              _______  (source: _______)  or: "not measured"
6. Open issues, our account                      _______  (source: _______)
7. Open issues, your account                     [blank until the meeting]
8. What changed for you, from your side          [blank until the meeting]
9. Next options: A / B / C / D — what each keeps, what each leaves unresolved
10. What we still need to decide, by when

Rows seven and eight stay blank in the deck you send. They are blank because only the customer can fill them, and pre-writing their reaction is a way of arguing with someone who hasn't spoken yet.

To see how the distinctions behave, consider an invented engagement. Everything in it is stipulated for illustration; none of it was observed.

A fictional staffing agency places licensed tradespeople through four branches. The signed scope promised to reduce last-minute shift cancellations caused by a placed worker's credential having lapsed. The service had two capabilities: a credential register tracking each worker's licences and expiry dates, and an eligibility check that flags, at scheduling time, which workers are blocked.

The delivery record says the register was configured in month one, piloted in one branch, and rolled out to all four by the end of month three. The eligibility check was built and switched on in month four. The integration into the agency's scheduling tool was scoped as a later phase and has not started, because the scheduling vendor's API requires a certification step and the customer-side sponsor who owned it changed roles in month six. The agency is also evaluating a replacement for that scheduling system, with a vendor decision expected in the next period.

The usage record says 1,180 of 1,240 placed workers have a record with current credential data. The expiry queue gets opened in three of four branches; the fourth keeps its own spreadsheet. The eligibility check exists as a manual report. Two of four branches run it, and nothing shows a shift decision changing because of it.

Did cancellations fall? Nobody knows. The scheduling system has a cancellation-reason field, but it is free text and no report reads it. That is the outcome row: not "improved," not "on track" — not measured.

3 — Include the work that did not go as planned

Material support issues belong in the deck, in your own account, before the customer has to raise them. Not every ticket — the ones that changed how someone worked.

In the invented engagement, two things qualify. In month two, eleven tickets about duplicate worker records after an HR export import; the cause was a field mapping, and it was fixed within the month. One branch still runs a manual deduplication pass every Monday, months later, because that is what the branch learned to do. In month five, expiry reminders stopped for six days after a configuration change. Thirty-four credentials expired inside that window. Whether the branches caught all thirty-four by hand is not recorded anywhere.

Notice what a ticket status does and does not tell you. The month-two issue is closed in the ticketing system. The behaviour it produced is still running. A closed ticket settles the supplier's queue; it does not settle the customer's concern. If your deck shows only closure rates, the branch manager with the Monday spreadsheet is going to spend the meeting deciding whether to mention it.

So the deck should carry four kinds of unfinished business:

• Support issues that changed the customer's work, with what caused them, what the effect was, and whether the workaround is still in place.
• Committed work that is not done, with a current status — including "scoped, not started" and "we recommend not doing this."
• Capabilities whose role changed. Something that made sense at signature may now be unnecessary, or may depend on a system that is going away.
• Your account, separate from theirs. State yours plainly and leave room for theirs. Do not write sentences that imply agreement you don't have.

4 — Compare what the next arrangement would justify

By this point the honest options are usually visible, and they are rarely just "renew" or "don't."

For the invented agency, four routes exist. Continue as it is: the register keeps working, the fourth branch keeps its spreadsheet, the eligibility check stays switched on without evidence that it changes anything, and the outcome stays unmeasured for another year. Continue with narrower scope: keep the register, park the eligibility check until the scheduling replacement is decided, and spend the next period on the fourth branch and on getting a measurement — either a report on the existing reason field or a thirty-day sample the customer agrees to log. Or wait: renew only the register, and revisit integration once the agency knows which scheduling system it will be on. A fourth possibility deserves a line too — the agency may conclude the register belongs with whoever wins the scheduling contract, and the right recommendation is to let it go.

Two rules keep this section honest.

A new capability promise cannot supply evidence for the previous period. If the integration would finally get built next year, that is next year's work. It does not retroactively prove that this year's promise was met, and presenting it as a highlight of the period is the move customers notice fastest.

Commercial terms stay separate from the factual account. Pricing, term length and scope belong on their own slide. If you also want to show financial value, label it as a scenario with stated assumptions — "if cancellations fall by this much, the register costs this much" — and only if the underlying number exists. In the invented engagement it does not, so there is no scenario. There is a question and a plan to answer it, which is a weaker slide and a much stronger position.

5 — Separate strategic review from the renewal decision

These are two conversations, and running them as one is where a lot of the distrust in this section comes from. GitLab's public customer-success handbook separates its executive business review from the renewal process, describing the review around the customer's organizational objectives and longer-term direction (Executive Business Reviews, GitLab handbook).

That is one provider's practice, and it is worth reading as context rather than as a validated template. It supports keeping the two conversations distinct. It does not supply your engagement records, and it does not establish that any particular meeting cadence improves renewal outcomes.

The practical version: decide which conversation you are having, and tell the customer. A strategic review asks where their business is going and what that implies about the work. A renewal conversation asks about terms, scope and the next period. They can happen the same week. They should not be merged without saying so.

If the review surfaces something that changes the commercial recommendation, take it seriously in the moment. "This changes what we'd suggest for next year. Can we put thirty minutes on the calendar Thursday to talk about scope?" is a sentence that keeps you in an honest relationship with the customer. The alternative — smiling through the discovery and producing a renewal number at the end — works once.

End this section by agreeing what is still missing. Who gets the cancellation reason data, by when, and what the next conversation will cover. A review that closes with clear next evidence is doing its job even if nothing was decided.

What the last slide should say

Close with a proposal proportionate to what you can actually support. Sometimes that is a renewal at the same scope. Often it is smaller than what the customer has today, because a piece of the service has stopped earning its place. Sometimes it is a short extension tied to getting the measurement nobody got last year.

All three of those are better than a proposal that requires the original pitch to have been true. The customer already knows what happened. The deck's job is to show that you know it too, that you can say it in the same room, and that the next year is built on the relationship as it currently stands rather than on the version you sold.

Frequently asked questions

What two things should a renewal presentation do, and in what order?

First, give an accurate account of the period already reviewed: what was agreed, delivered, used, observed, and what remains open. Second, propose something for the coming period, which may be continuation, narrower scope, a different service, or stopping part of it. The second part must not rewrite the first.

How should delivered, adopted, and useful be distinguished?

Keep them as three separate statements with different next-period responses. Delivered means built, configured, and shipped, drawn from delivery notes, change logs, or tickets. Adopted means people used it, drawn from usage logs, but logs count events, not decisions. Useful means an outcome moved, usually shown by the customer's systems, the customer's judgment, or nothing at all.

What should be done when a capability was used but no one agreed how success would be known?

Name it as an unmeasured engagement rather than a success with thin evidence. The honest move is to propose how to measure it, since the capability's use alone does not establish that it worked.

Why should some rows stay blank in the deck sent to the customer?

Rows for the customer's account of open issues and what changed on their side stay blank because only the customer can fill them. Pre-writing their reaction is a way of arguing with someone who has not spoken yet.

How should commercial terms and new capability promises be handled?

Commercial terms such as pricing, term length, and scope belong on their own slide, separate from the factual account. A new capability promise cannot supply evidence for the previous period. If financial value is shown, label it as a scenario with stated assumptions, and only if the underlying number exists.

Further reading

• Choose a Customer Story That Answers This Prospect’s Question — Read when a renewal slide uses a customer quote as proof; it separates experience from measured outcome and treats approval to quote separately.
• Two Valid Metrics Tell Opposite Stories. What Should the Deck Lead With? — Use when usage and outcome measures point in different directions; it helps separate definition drift from real behavior change and keeps the counterpoint visible.