Playbook

How to Find a Collaboration Partner: The ICP × JTBD × Value-Chain Framework

July 23, 20268 min read

Short answer: the right collaboration partner passes three filters. They share your ICP (you sell to the same customer), they serve the same JTBD (the same job that customer is trying to get done), and they fill a different slot in the value chain (they offer something you don't — so they complement you instead of competing). Same buyer, same goal, different part of the work. Get those three right and the partnership almost designs itself.

Most founders search for partners by vibe — someone they met at an event, a founder they admire, a company with a nice logo. That's how you end up with a "partnership" that produces one joint webinar and then dies. A real collaboration partner isn't found by chemistry. It's found by fit. Below is the framework to find one by hand, and then the faster way to let AI do it for you.

Filter 1 — same ICP: you're chasing the same buyer

Your ideal customer profile is the specific buyer your business is built to serve — their size, sector, budget, and moment of need. The first filter is simple: your partner must be selling to the same ICP. Not a similar one. The same one.

Why it's non-negotiable: a collaboration only compounds if one plus one lands in front of one shared customer. If you sell to seed-stage SaaS founders and your "partner" sells to enterprise procurement teams, there's no shared customer to bundle for — you're just two businesses that happen to like each other. Write your ICP down in one sentence before you evaluate anyone. If a candidate's customer doesn't match that sentence, stop there.

Filter 2 — same JTBD: the deeper job the customer wants done

People don't buy products; they hire them to make progress on a job. The job-to-be-done is the outcome beneath the purchase. A founder doesn't want "a landing page" — they want to launch and get their first customers. They don't want "an accountant" — they want to stop worrying about compliance.

Your partner should be serving that same underlying job from a different angle. When two offers point at the same JTBD, a bundle feels like one solution to the customer rather than two vendors stapled together. That coherence is what makes the combined pitch convert. If your candidate technically shares your ICP but is solving a completely unrelated job, the bundle will feel arbitrary — and customers can smell arbitrary.

Filter 3 — a different value-chain slot: complement, don't compete

Here's the filter founders get backwards. It's tempting to look for a partner who does what you do, only bigger or in another region. That's not a partner — that's a competitor. The third filter is the hard rule: your partner must not offer what you offer. They fill a different slot in the value chain, so together you cover more of the customer's journey than either could alone.

Think of the value your shared customer moves through as a set of slots. Five are core-facing:

Product Build. Making the thing — engineering, design, the core product or deliverable itself.
Go-to-Market. Getting attention — marketing, content, demand generation, brand, launch.
Revenue & Customers. Closing and onboarding — sales, distribution, the channel that turns interest into paying customers.
Retention & Scale. Keeping and growing accounts — success, support, expansion, the machinery that scales what's working.
Ecosystem & Expansion. New surfaces — integrations, marketplaces, new segments and geographies that extend the reach of the core.

And three are supporting slots that make the core possible:

Operations. The back office — finance, legal, compliance, the plumbing that lets the business run.
Expertise / Advisory. Specialist know-how — the consultant, the domain expert, the fractional leader the customer can't hire full-time.
Data / Insights. The intelligence layer — analytics, research, the data that tells everyone else what to do next.

Map your own slot first. If you're a design studio, you sit in Product Build. Your ideal partners live in the adjacent slots — a developer to ship what you design (Product Build's neighbor toward delivery), an SEO/AEO specialist in Go-to-Market to make it visible, an agency in Revenue & Customers with the audience you lack. The moment a candidate's slot is your slot, you've found a competitor, not a collaborator.

The two archetypes a partner falls into

When all three filters pass, the collaboration takes one of two shapes:

Bundled front-end offer. Two complementary products sold together to one shared customer. Neither product changes; they're packaged as a joint offer where buying both solves the whole job. This is the classic scenario collaboration — silk garments plus in-store cleaning, an app plus the audience that promotes it.
Complementary back-end build. Several capabilities combined so the client buys "one thing" that multiple businesses actually deliver. The customer sees a single deliverable — a website, a launch, a full campaign — while a designer, a developer, and a marketer each own their slot behind the scenes.

The hands-on process

You can run this today, no tools required:

Write your ICP in one sentence. The exact customer you both need to share.
Name the JTBD. The outcome that customer is really buying — not your feature, their progress.
Map your value-chain slot. Which of the eight slots do you occupy? Be honest and specific.
List the adjacent slots. Which neighboring slots does your customer need filled that you don't fill?
Find startups in those slots. Search, ask your network, comb through communities for companies serving your ICP in an adjacent slot.
Sanity-check for overlap. If they do any part of what you do, drop them — capability overlap turns partnership into competition.
Reach out with the shared customer, not your pitch. Lead with the buyer and job you both serve; propose the bundle.

Done well, this works. Done by hand, it's slow — you're manually holding three filters in your head while searching a market you can only partially see, and most founders give up somewhere around step five.

The framework, automated

This is exactly what Ordana's AI Find Collaborators does for you. It already holds your ICP, your JTBD, and your value-chain slot as context — then runs the search you'd do by hand across a catalog of existing Ordana startups and the open web, using Gemini plus Exa deep search. It applies the same three filters automatically: same buyer, same job, different slot, no capability overlap.

From there it proposes the fit and routes it honestly by the revenue-source test — a single unified journey with two shapes. One revenue source becomes a 1-on-1 project — a lever to grow now and pay from the upside, with owner-led governance. Two or more revenue sources become a multi-party scenario — a flywheel where every sale carries a built-in upsell, turning each partner's customers into each other's, with vote-based governance. No scenario bias — the AI picks the shape that actually fits.

Then it does the tedious part: one-click create the collaboration, send personalized invites (each invitee sees an AI-estimated earning potential — always an estimate, never a promise), auto-generate the contract, and stand up the automatic revenue split. Ordana takes a flat 5% off the top; the collaborators' shares divide the remaining 95%. Contract plus auto rev-share, live in about fifteen minutes.

The takeaway

Finding a collaboration partner isn't luck and it isn't networking. It's a filter: same ICP, same JTBD, different value-chain slot. Run it by hand and you'll find better partners than most founders ever do. Let Ordana's AI run it and you'll find them in an afternoon — with the contract and revenue share already built.

Related reading:

Find your collaborators on Ordana → Free to join — pay only when revenue flows.