1. Hours of persuasion per deal: from first contact to signature, how many hours did you actually spend? Past five or six, stop expecting a partner to replicate it — he has neither the patience nor your background.
2. What a partner earns per deal: your margin × his rate, divided by the hours in step 1, is his effective hourly rate. Below what else is on his desk, no percentage will move him.
3. The answer is rarely either/or: keep the persuading, give away the moving. He sources leads, makes first contact, handles invoicing and front-line support; you take the meeting where the decision is made. Split commission the same way — a small rate for a referred lead, a full one for a closed deal.
Turning down a bad channel: "The customers you can bring aren't the ones we're set up to serve, and rushing it would let both sides down. If you ever have the XX type, come to me and we'll do it as a referral."
Mechanism: whether to do a thing in-house or buy it outside comes down to which side carries the higher transaction cost. When the work depends on highly specific knowledge and quality is hard to measure, the market is expensive and doing it yourself is cheap. For standardised, measurable work it's the reverse.
Boundary: this framework assumes you can calculate margin per customer and hours of persuasion. Early on, with a handful of deals, nothing lands in the right box — go and sell a few more yourself and learn the process. A channel amplifies certainty, not hope.
Mechanism: an agent's goals are structurally different from yours — he optimises his own total income, not your volume — and his constraint is time, not willingness. So there are only two ways to make him sell more: raise his output per hour, or cut his cost per hour. The second is usually cheaper, because it spends your marginal time rather than your margin.
Boundary: this arithmetic assumes a rational businessperson, which independent agents, resellers and referrers broadly are. If the person is an employee inside a large company (a systems integrator's pre-sales, say), his ranking also contains internal targets and departmental politics — there, what you influence is how he's measured, not his commission.
Situation: you've closed an order on a marketplace, content platform, app store or services site, and you're delivering.
"It's on its way. There's one thing that won't fit on the platform — I've put together a {pre-install checklist / sizing chart / answers to the common questions}. Message me at {contact} and I'll send it straight over, and you can ask me directly later instead of waiting on a support ticket."
Why it works: you're giving a reason, not making a request. "Add me for discounts later" asks him for something; "there's a thing that won't fit on the platform" gives him something — and actually gives it.
Red line: don't route around the platform's payment and transaction rules. You're pulling the future relationship over, not stealing this order; the penalty for the latter is usually a closed shop, and they can detect it. Playing the inside of the rules hard is enough.
Mechanism: a platform doesn't price by allocating cost. It subsidises the side that's harder to attract and more price-sensitive, and charges the side that's more locked in and more willing to pay. You start out as the scarce, subsidised side; once supply like yours is plentiful you've moved to the side that gets charged. Withdrawing the subsidy isn't betrayal, it's repricing.
Boundary: not every platform ends up extracting heavily — regulation, competition and what stage the platform is at all change the pace. This isn't a prediction, it's risk management: arranging your customer assets on the assumption that rates will rise someday costs very little; arranging them on the assumption it will always be good to you costs the whole business.
1. How the map divides: which customer segment belongs to partners, which to direct, and which named accounts nobody touches. The list must be published in advance — declaring an account strategic after the fact admits the rules are yours to rewrite.
2. Deal registration: the partner submits a named account, you confirm within two working days, and it's his for 90 days regardless of who ends up negotiating. First come, first served — and your direct team runs the same process.
3. One price: same customer, same configuration, same price whichever path he buys through. Discount into the partner's margin, never into the customer's quote.
4. Who arbitrates: name the person and the number of days to a decision. The value of a rule isn't being right every time, it's that he knows there's a ruling and it comes fast.
When it does collide: "By the registration record this one is yours — direct is standing down, and I'll put A on it for technical support." The cost is one deal; what it buys is the credibility of the whole system.
Mechanism: people judge a split not only by their own share but by whether the split itself is legitimate, and once they read it as exploitation they resist hard enough to give up their own gain. So the first job of a rule isn't to decide correctly, it's to make a partner believe in advance that what he invests won't be taken. Without that belief his optimal move is to invest nothing.
Boundary and red line: lines exist to make the rules predictable, not to lock a partner in. Don't turn "protecting the channel" into "the customer may only buy the way I say" — restricting his choice eventually costs you close rate. It holds for individuals too: agreeing what counts as your introduction, for how long, and at what fee, before the fact, is far less damaging to a friendship than splitting it afterwards.
Ninety minutes on whatever you're selling right now.