The partner renewal playbook: deepen, hold, or sunset
How to renew a technology partnership on evidence: the conversation, the commercial terms, and the relationship care that makes deepen, hold, or sunset a clean call.
Renewal is the point in the lifecycle where you stop inheriting last year's story. A strategic alliance that is not re-chosen becomes furniture. Tech partnerships rot the same way: the paper auto-renews, the listing stays up, and nobody can say whether you would start this partner again.
A renewal is a decision, not a ceremony. Deepen, hold, or sunset. It should rest on evidence both sides can see, a conversation that happens before the term date, and commercial terms that match the next year rather than the first pitch. Relationship care matters because you may need this person later, at a different company, with a better product fit. Care is not the same as keeping a dead integration live.
This playbook is the evidence pack, the conversation, the term changes worth making, and how to leave the relationship in a state you would not be ashamed of.
The 60-second version
If you only read one section, read this one:
- Renewal is deepen, hold, or sunset. "See how it goes" is not a decision.
- Build an evidence pack first. Adoption, pipeline, reliability, owners, ICP fit. No adjectives until the table is full.
- Talk before the term date, in a QBR shape, with the decision on the agenda.
- Commercial terms should match the next motion, not the original hope. Do not pay co-sell rates for a quiet referral.
- Hold is a first-class outcome. Keep it running, stop investing, write the review date.
- Sunset with notice, a customer path, and a written close. Silence is how you burn the next cycle.
- Care is specific. Tell them early, thank the humans, do not surprise their sellers or their customers.
Evidence before conversation
If you cannot fill this table, you are not ready to renew. Instrument, then book the meeting. Borrow the panel from partnership metrics.
| Evidence | What you need | How it points |
|---|---|---|
| Adoption | Weekly actives, active-to-install, trend over 2 to 4 quarters | Deepen if real and rising; hold if low but used; sunset if near zero |
| Pipeline | Sourced and influenced, with definitions you both use | Deepen if a line you would fund; sunset if none and no story |
| Reliability | Error rate, tickets, incidents, time to patch | Hold or sunset if you are a support tax |
| Owners | Named humans still in seat, reply times | No owner is a sunset or a pause until one appears |
| Fit | Current partner ICP, overlap, competitive drift | Drift without pipeline is a sunset candidate |
| Cost | Eng hours, marketplace fees, legal, MDF | Cost without outcome fails the keep test |
| Paper | Term date, auto-renew, exclusivity, data, deprecation notice | Know what happens if you do nothing |
Do nothing is a choice. Auto-renew on a zombie is how programs fill with furniture. Put the term date on a calendar 90 days out. Business relationship management is mostly this: notice the date, bring facts, decide. Re-score as if they were inbound today using your prioritization framework. A partner who would not earn a deep slot now should not get one because you already built it.
Write a one-page internal memo: recommended path, the numbers that force it, what you will ask, what you will not reopen. Share it with product and, if money moves, finance. Do not discover your own disagreement in front of the partner.
The conversation
Put the decision on the agenda. "Renewal: deepen, hold, or wind down" is a line item, not a surprise at minute 55. Use QBR structure: short scoreboard, diagnosis, then the call.
Lead with what is true. Then the recommendation. Then what you want from them if they want a different path. "Adoption is flat for two quarters, no sourced pipeline, we recommend hold: keep invoice sync, no new objects, review in six months." That is a conversation. "We love this partnership and also we need to be more strategic" is not.
| Path | What you say | What you ask them to do |
|---|---|---|
| Deepen | The numbers support a bigger bet | Name capacity for the expansion, agree one bet |
| Hold | The numbers support keeping, not investing | Accept no new scope, keep an owner for incidents |
| Sunset | The numbers do not support keeping it live | Agree dates, customer path, listing takedown |
If they want deepen and you want hold, do not split the difference into a fake project. Offer the evidence they would need by the next review date. If they want to keep a listing you consider a zombie, you can still sunset your investment and leave a freeze for remaining users.
Commercial talk belongs after the path is clear. Terms are how you fund the path, not how you avoid choosing one. Bring the agreements points that actually change: term length, auto-renew, rev share or referral fee, exclusivity, deprecation notice, data. Leave the rest.
Relationship care in the room: thank the people who did work, even if the partnership did not hit. Do not litigate old incidents unless they are still open. Do not copy their boss as a power move in the first meeting. If you need an escalate, say you will, then do it with facts.
Follow with a written decision the same week. Path, dates, term changes, who tells customers and sellers. This is the renewal artifact. Without it, people will remember the tone and forget the call.
Commercial terms and the three paths
Deepen. You are buying more of a working thing: extra scope, a co-sell motion, a region, maybe better incentives because there is now pipeline to attach them to. Shorten the next review if the expansion is a test. Do not grant exclusivity as a thank-you. Exclusivity is a price they pay for with volume, not with history.
Hold. Freeze new build. Keep production up. Maybe cut MDF and co-marketing. Maybe move them down a tier so the program stays honest. Auto-renew can be dangerous here; a 12-month hold with a review date is cleaner than a rolling term nobody will look at.
Sunset. Stop new investment, publish customer-facing and hard-off dates, take listings down in order, migrate or export, close in writing. Details live in the sunset guide. On paper: end the term, survive the clauses you need for customer wind-down, do not pick a fight on leftover MDF if it costs more in time than in cash.
| Term | Deepen | Hold | Sunset |
|---|---|---|---|
| Duration | 12 months with a mid-point review | 6 to 12 months, explicit | Wind-down schedule, then end |
| Economics | Match the new motion | Match the remaining motion, often lower | None going forward |
| Exclusivity | Rare, only with volume | No | No |
| Listing | Expand or refresh | Stay, marked as maintained not promoted | Takedown plan |
| Support | Keep or improve SLAs | Best effort, documented | Until hard-off only |
Finance should see the economics before you promise them. Partnerships should not freelance discounts in the QBR.
If the partner is also a vendor you pay, split the conversations: product contract versus alliance. Mixing them is how a renewal becomes a hostage situation.
Relationship care when the answer is hard
You can sunset a connector without humiliating the people. Tell the DRI first, not Twitter. Give them a chance to bring a last fact. Then decide.
Tell your own sellers the same week so they stop pitching a dying motion. Tell remaining customers with a path. Copy the partner on customer language so you do not contradict each other.
Leave a door that is real or leave none. "If you hire an owner and adoption returns, we will re-score in H2" is a door. "Let's stay close" is not. People who write for First Round Review tend to prefer the first kind of sentence. You should too.
Do not use renewal as a dump of every frustration from two years. Pick the facts that change the path. Save the rest for your internal memo so you do not repeat the same bet.
If you deepen, staff what you just promised. A warm renewal and an empty sprint is a stall by spring.
Common mistakes, and the fix
Auto-renew by inattention. The fix: term dates 90 days out, evidence pack, a written path.
Renewal as a dinner. The fix: QBR with a scoreboard. Dinner after, if you still like each other.
Paying more to avoid a hard conversation. The fix: incentives follow a motion. They do not buy courage.
Sunsetting in silence. The fix: meeting, dates, customer path, close note. See the sunset checklist.
Treating hold as a failure, so you fake a deepen. The fix: hold in public on the program. Empty tiers and fake expansions cost more.
Reopening the whole contract. The fix: change the two or three terms that match the path.
FAQ
When should we start the renewal process? Ninety days before the term date, or at the QBR closest to the anniversary if there is no hard date. Earlier if adoption has been dead for two quarters.
What if there is no contract, only a listing? You still renew the relationship. Decide whether the listing stays, whether anyone owns it, and whether you will spend hours. Paper is not a requirement for a decision.
How is hold different from freeze? Same idea in practice: no new investment, keep it running for remaining users, review date written. Use whatever word your team will actually apply.
Do we need a lawyer for every renewal? No. You need one when terms, data, exclusivity, or wind-down dates change. The path decision is commercial and operational first.
What if they refuse sunset and we still want out? Follow the contract's termination and customer clauses. Be decent with users. You do not need their blessing to stop investing, and you may need counsel to stop the paper.
Should renewal be in our OKRs? The outcomes should: live partners that meet a health bar, slots freed, expansions that hit a leading indicator. "Renew five partners" is activity. See partnership OKRs.
Can we deepen one motion and sunset another with the same company? Yes. Example: keep the integration, kill an unused co-sell experiment. Write both so sellers do not mix them.
What do we tell the board? How many you deepened, held, or sunset, and what capacity that frees. A logo count that never goes down is not a renewal process.
Further reading
- Strategic alliance for why alliances that are not re-chosen fade.
- Business relationship management for owning the relationship through a decision.
- First Round Review for operator-plain language when the answer is hard.
- When to sunset a partnership for the exit project.
- Partner QBR for the meeting shape.
- SaaS partnership agreements for the terms that actually matter.
The short version
Renew on evidence. Fill the scoreboard, re-score as if they were new, and pick deepen, hold, or sunset before the term date. Say it in a QBR, write it the same week, and change only the commercial terms that match the path. Hold is allowed. Sunset with notice and a customer path. Do not auto-renew furniture, and do not buy your way out of the conversation with extra incentives. Care for the humans. Staff what you just promised.
If you want a deepen / hold / sunset pass on the current portfolio, with the evidence pack and the term dates, that is what a Partner Audit is for.