The first 90 days of a technology partnership
A week-by-week plan from signed intent to a scoped build or first referral, so the first 90 days of a technology partnership produce work, not a quiet Slack channel.
The signature, or even the clear verbal yes, starts a clock. For about 90 days both sides still have attention. After that, other roadmap items win unless the partnership is already real: a scoped build in motion, a live listing you will maintain, or a first referral with a named owner. Most partnerships fail in the gap between yes and the first visible work, not in month nine.
This is not the same as partner onboarding, which is how you enable a partner to sell. The first 90 days cover the whole loop from intent through the first motion: kickoff, scope, access, a thin slice of build or GTM, a public story only when it is true. Go-to-market starts here as a choice, not as a launch party.
This post is a week-by-week plan you can run with a small team. Adapt the build-heavy weeks if your first motion is referral only. Do not skip the owner, the date, or the definition of done.
The 60-second version
If you only read one section, read this one:
- 90 days exist to produce one real motion, not a stack of intros. Scoped build underway, or first referral path live, with owners.
- Weeks 1 to 2 are kickoff, owners, and the written slice. If this slips, the rest is fiction.
- Weeks 3 to 6 are access, scope freeze, and work on the slice. Stop adding objects.
- Weeks 7 to 10 are a thin launch or a first joint action. Internal first, public second.
- Weeks 11 to 13 are the first review. Keep, adjust, or pause. Book the QBR window.
- One DRI on each side the entire time. A channel is not an owner.
- Do not announce a date you have not staffed. Public is a week 8 to 10 question, not a week 1 treat.
What "done" looks like on day 90
Pick one primary finish line before week 1, based on the motion you chose in referral vs reseller vs co-sell.
| First motion | Day-90 done | Not done |
|---|---|---|
| Integration | Scope frozen, build in progress or live in beta with a named customer | Slack, a deck, "API access requested" |
| Referral | Path live: how they send, how you accept, one real intro or a dated first campaign | A PDF of "how we partner" |
| Marketplace listing | Listing submitted or live, support ready, tracking in place | Copy in a doc nobody filed |
| Co-sell (only if the product story already exists) | JVP, kit v1, one mapped account list, deal-registration rule | A promise that sellers will "keep us in mind" |
If you try to finish all four, you will finish none. The prioritization logic still applies inside a single partnership: one deep slice.
Day 90 is also a health check. Owners still answer. The working doc is current. Shared customers, if any, know what is coming or are already in a beta. Legal is either closed or explicitly parked with a list of what you will not do until it closes. That is enough to enter the long lifecycle stage of partner success. It is not enough to declare victory.
Write the finish line in the kickoff note. If you cannot, you are not in day 1 of a partnership. You are still in discovery.
Weeks 1 to 4: owners, slice, access
Week 1. Run the kickoff. Two DRIs, first-slice in and out, comms path, next holds. Send the note the same day. Create the working doc. If you cannot name an engineer (for a build) or a seller lead (for referral), pause. Do not start a 90-day theater.
Week 2. Freeze the slice on paper. For a build: objects, sync direction, auth, outs, success test. For referral: ICP overlap, intro template, what a qualified intro is. Draft the joint value proposition in one paragraph. Start enablement as a kit outline, not a training day.
Week 3. Access. Sandbox, keys, marketplace accounts, CRM partner field, ticket tag. This week is boring and it is where delays hide. Put every access item in a table with an owner and a date.
Week 4. Internal checkpoint. Slice still true? Capacity still reserved? If the slice grew, cut it back. If the partner went quiet, treat it as an early stall.
| Week | Your side | Their side | Artifact |
|---|---|---|---|
| 1 | Kickoff, note, working doc | DRI in the room, calendar holds | Decision note |
| 2 | Scope or referral rules drafted | Comments, named reviewer | Frozen slice v1 |
| 3 | Access requested and chased | Access granted | Access table |
| 4 | Checkpoint, cut scope if needed | Confirm capacity | Go / pause |
Onboarding tasks (kit, product orientation) run in parallel in these weeks so their sellers are not waiting on the build to learn the sentence they will say.
Weeks 5 to 10: the slice, then a thin launch
Weeks 5 to 7. Do the work. Engineers build the first slice or you stand up the referral path and register it in the CRM. Meet weekly for 15 minutes on blockers only. No new objects. No extra regions. If a customer wants more, log it for after day 90.
Beta users belong here if you are building. One or two shared customers, in writing, with a support path. A beta with nobody named is a private demo.
Week 8. Internal readiness. Support has a runbook. Sales on both sides can say the JVP. Tracking exists: partner field, listing UTM, or a simple install count. Metrics you will actually look at, not a dashboard project.
Weeks 9 to 10. Thin launch. Listing live or submitted, integration in production for beta, or first referral campaign in the wild. Public announcement only if the thing a customer would click is real. Co-marketing is optional and small: a blog, a marketplace post, a note to the two CS teams. Save the webinar.
| Week | Build motion | Referral motion |
|---|---|---|
| 5 to 7 | Implement and test the slice | Enable a small seller group, test one intro |
| 8 | Support runbook, tracking | CRM path proven with a dummy or live intro |
| 9 to 10 | Beta in prod, listing if ready | First real intro or dated sequence live |
If week 8 shows you will miss a public date, move the date. Do not launch a broken slice to keep a tweet. Reseller and full co-sell rarely belong in the first 90. Default to referral or a thin integration. You can expand later.
Weeks 11 to 13: first review, then the long stage
Week 11. Collect the facts. Did we hit the day-90 done line. What slipped. Adoption or intro quality, even if the numbers are small. Tickets. Owner health.
Week 12. Joint review, 45 minutes. This is a baby QBR, not a retrospective of feelings. Scoreboard, what to finish, what to cut, whether the next quarter is deepen, hold, or pause. Write keep / adjust / pause.
Week 13. Handoff into partner success. Name who owns the weekly health look. Put the first real QBR on the calendar 60 to 90 days out. Update the kit with what you learned. Close leftover access issues. If you are pausing, say so to both sales teams.
| Review question | Keep | Adjust | Pause |
|---|---|---|---|
| Slice shipped or path live? | Yes | Almost, with a named miss | No, and no date |
| Owners still real? | Yes | New name needed | No owner |
| Any customer or intro signal? | Yes, even small | Not yet, but beta booked | None, and no plan |
| Worth another slot next quarter? | Yes, listed bet | Yes, smaller bet | No |
Partnership OKRs for this window should be leading: time to first live, first qualified intro, beta customer using the slice. Sourced revenue by day 90 is luck, not a plan.
If day 90 is a miss because you never kicked off, reset with a new week 1 or stop. An endless "first 90" is a stall with a nicer name.
Common mistakes, and the fix
Treating 90 days as relationship time. The fix: a finish line in week 1, checked in week 4 and week 12.
Scope that grows every Friday. The fix: freeze in week 2. New ideas go on a list for after the first slice.
Announcement in week 1. The fix: public after something a customer can use. Intent tweets are optional and dated as intent.
No access table. The fix: week 3 is access. Keys, sandboxes, marketplace logins, CRM fields. Chase like a project.
Skipping support. The fix: a runbook in week 8. Shared customers will find the bugs before your tweet does.
Two motions at once. The fix: integration or referral as primary. Co-sell later, when there is a sentence and a kit.
FAQ
Does the clock start at signature or at verbal yes? At the yes you are willing to spend hours on. If paper is slow, run non-production weeks anyway and say so. Do not wait three months for a PDF to begin week 1.
What if engineering cannot start until week 6? Compress weeks 1 to 4, be honest in the kickoff, and redefine day 90 as "scope frozen, slot dated, kit ready," not "live." A fake live date helps no one.
How does this change for a referral-only partner? Less engineering, same owners and dates. Weeks 5 to 7 are enablement and CRM path. Done is a live intro path, not a connector.
Should we include a customer in the kickoff? Usually no. Bring a named beta customer in weeks 5 to 7 if they have agreed. Kickoff is for the two companies to assign work.
What if the partner is a large platform with a slow app review? Put review on the timeline as a dependency. Day 90 may be "submitted and in review" rather than "listed." That is still a real finish line. Waiting without submitting is not.
Can one person run 90 days for three partners? Not if all three are in weeks 5 to 7. Stagger starts. One partner in build, others in earlier weeks or waitlist.
When do we hire to support this? After you have run this plan once yourself and the motion is real. See when to hire partnerships. Do not hire to replace a plan you never wrote.
What do we tell the board at day 90? The finish line you set, whether you hit it, adoption or intros, and the next bet. Not "great relationship."
Further reading
- Go-to-market for the motion you are actually standing up.
- Partner onboarding for enablement inside this window.
- The SaaS partnership lifecycle for what 90 days sit inside.
- Partner enablement kit for what to ship in weeks 2 to 8.
- Partnership metrics for what to look at in week 12.
The short version
The first 90 days turn a yes into one real motion. Weeks 1 to 4: kickoff, frozen slice, access, a go or pause. Weeks 5 to 10: build or stand up the path, get support ready, launch thin. Weeks 11 to 13: review with numbers, then hand off to partner success or pause. One DRI each side, no scope creep, no public date without a staffed slice. If you miss because you never started, reset or stop. Do not extend forever.
If you want a 90-day plan on a specific partner, with the slice, owners, and finish line written down, that is what a Partner Audit is for.