Resources · Playbook · 9 min read

How to start a partner programme from nothing

The first ninety days, in the order that actually works: pick one partner type, sign five, get one deal, then build the programme around what you learned.

Do not build a programme first

The instinct is to design tiers, write a partner agreement, choose a PRM and build a portal, then go and find partners to put in it. This is backwards, and it is why most partner programmes are quietly shut down within a year.

You do not yet know which kind of partner works for your product, what they will need, or what commission makes the maths work for them. Every hour spent on programme design before your first partner-sourced deal is an hour spent guessing.

Sign five partners before you design anything. The programme is the thing you write down afterwards.

Pick exactly one partner type

Referral partners are almost always the right place to start. They need the least from you, they cost nothing until a deal closes, and they tell you fastest whether partner-sourced revenue is real for your product.

Resellers and agencies are more valuable and much slower. They need margin certainty, deal protection and product depth before they will invest a sales cycle. That is a reasonable thing to build for partner six through twenty, not for partner one.

Find the five who are already doing it

Your first partners are not strangers. They are consultants who already recommend you, agencies who already implement you, and customers who already refer you and have never been paid for it.

Search your CRM for deals where the source is a person rather than a campaign. Ask your five happiest customers who helped them choose you. That list is your programme.

  • Consultants who appear repeatedly in won-deal notes.
  • Agencies your customers mention when you ask who set things up.
  • Adjacent vendors whose sales teams keep running into you.
  • Customers who have referred someone unprompted.

Make registering a deal take twenty seconds

This is the single highest-leverage decision in the first ninety days, and it is the one most programmes get wrong by buying a portal.

Every step between a partner having a name and you having a record loses you a percentage of registrations — permanently, and invisibly, because you never learn about the ones that did not happen. If registering is harder than sending an email, they will send the email, and the deal will land in your CRM as inbound with no partner attached.

If your partner has to remember a password to register a deal, you have designed a system that measures how motivated they are, not how much revenue they bring.

Decide the commission before you need it

Pick something simple and defensible: a percentage of first-year collected revenue is right for most products. Write down the protection window, the clawback rule and when payment happens.

The specific number matters less than never changing it retroactively. A partner who discovers their rate moved after they did the work does not complain — they stop sending deals.

DecisionA sensible defaultWhy
Rate15% of collected revenueHigh enough to be worth the introduction, low enough to survive a CFO
BasisCollected, not bookedYou cannot pay commission out of an invoice nobody paid
Protection90 days from registrationLong enough for a real cycle, short enough to stop squatting
Clawback60 daysCovers early churn without punishing the partner for your retention
PaymentMonthly, unpromptedChasing an invoice costs more than the commission is worth

Measure two things, and only two

Activation rate — the share of signed partners who have ever registered a deal. And time to first registration. Everything else is a distraction until those two are healthy.

A programme with forty partners and four active ones does not have a recruitment problem. It has an activation problem, and signing twenty more partners will make it worse.

What ninety days should look like

Paying the first commission fast and slightly early is worth more than any enablement asset you could build. It is the only proof that matters, and partners tell each other.

WeeksThe one thingDone when
1–2List everyone already referring youYou have twenty names and can rank them
3–4Sign five, with one written commission ruleFive people know they will be paid, and how
5–8Make registration frictionless, then ask for dealsFirst registration arrives without you chasing it
9–12Pay the first commission early and visiblyThe first partner has been paid, and the others know

MacroBP does the part that is hard to do by hand

Registration in twenty seconds without an account, revenue reconciled to what actually collected, and commissions that settle themselves.

Get started →