The onboarding checklist closes in about three weeks. The working shape of the relationship is set over the same ninety days, by default, and it outlasts the contract.
Onboarding a distribution partner has a visible half and an invisible half.
The visible half is a list. Contract signed, first order placed, product data sent, logo files shared, a quarterly call put in the calendar. It takes about three weeks and it gets finished, because every item on it has an owner and a tick box. When the last box is ticked, both sides feel that the relationship has been set up.
The invisible half runs over the same ninety days and nothing on it is written down anywhere. Who reports what, in whose format. Which of their people you actually end up talking to. The size of the first order and the assumption underneath it. Which retailers hear about you first, and in what order. Whether anybody is writing down what they see at day sixty.
Three years later, it is the invisible half that describes the relationship. Problems arrive through one person. The forecast comes in their format, late, and nobody can reconcile it with yours. The accounts that matter were opened in an order you did not choose. And when you raise any of it, the answer is some version of this is how we have always done it with you.
Nothing on the checklist caused that. The checklist was correct and finished. The point of this piece is that the two halves run in parallel, they are set by different mechanisms, and only one of them is still negotiable after day ninety.
Onboarding is not a document transfer. It is default-setting: the window in which the working shape of the relationship is established, before either side has any history to appeal to. The checklist closes in three weeks. The defaults outlast the contract.
Why the first ninety days are different from the next thirty months
There is a narrow window in which you can ask for almost anything without it being a renegotiation.
Before a partner has shipped, invoiced, hired, or taken shelf space for you, every request is a term. After they have, the same request is a change. The words are identical; the cost is not. Asking for monthly sell-out data by retailer in week two is an administrative detail. Asking for it in month eighteen is an accusation — it implies you have stopped believing their numbers, and they will hear it that way whether or not you meant it.
This is not psychology. It is the economics of specific investment. Once a partner has put money into something that only has value inside this relationship — a dedicated salesperson, registered product dossiers, shelf space negotiated in your name, warehouse stock — the balance of bargaining power moves, because that investment cannot be redeployed. Klein, Crawford and Alchian named this the hold-up problem (the risk that, once one side has sunk an asset specific to the relationship, the other can extract value from it after the fact). Williamson called the underlying property asset specificity — the degree to which an investment loses value outside its original use.
The standard reading of hold-up is that it protects the partner from us. It also runs the other way, and that direction is the one brands miss. After month six, our ability to set terms has fallen, because the market now reaches us only through them.
What follows from that: everything you intend to have for three years must be installed while it is still a term. The ninety-day window is not a courtesy period. It is the only period in which governance is free.
Default one: which direction information flows, and in whose format
The checklist asks whether a reporting obligation exists. It does, in almost every agreement — "the Distributor shall provide monthly sales reports."
What the checklist never asks is the only question that matters: sales of what, to whom, measured when, in whose format.
If this is not fixed in the first ninety days, it fixes itself, and it always fixes itself the same way: the partner sends what their own system already produces. That is almost always sell-in — what they bought from you — because that is what their accounting needs. It is almost never sell-out — what left their warehouse to a retailer or a consumer — because producing that requires them to pull data from someone else.
The gap between those two numbers is the entire subject of The Month-Seven Test. A partner can report perfect numbers for two years while your product sits in their warehouse.
The expensive part is not the data. It is the format. Once eighteen months of reports exist in their shape, asking for a different shape means asking them to rebuild a process, and the request will be refused politely and indefinitely — not out of bad faith, but because the person who would have to do it has other work.
Install in the first ninety days:
- the unit of measurement (sell-out, by retailer or channel, monthly)
- the format, as a file you supply — not a specification they interpret
- the date it arrives, and what happens when it does not
- one worked example, filled in together, before the first real one is due
The last item is the one that is always skipped and always decides the rest. A template nobody has filled in together is a template that will come back wrong twice and then come back in their format.
Default two: which of their people actually touch the brand
You negotiated with their principal. You will not work with their principal.
The onboarding checklist records a named contact, usually one, usually senior, usually the person who signed. What it does not record is the set of people who will decide, week by week, whether your product gets pushed: the account manager who walks into the retailer, the person who builds the promotion plan, the one who allocates warehouse space among eleven brands.
If those people are not identified in the first ninety days, you get a single-channel relationship — everything you know about the market arrives through one person, and everything the market knows about you leaves through that same person. That is comfortable for exactly as long as that person stays and stays interested.
Johanson and Vahlne described the cost of being on the outside of a network as liability of outsidership — the disadvantage of not being embedded in the relationships through which a market actually operates. Appointing a partner does not remove that liability. It delegates it, and leaves you unable to see whether it is being removed or merely hidden.
Dyer and Singh, looking at the other side of the same thing, found that the returns in a partnership come disproportionately from knowledge-sharing routines — repeated, structured exchanges, not goodwill. Routines are made of named people and recurring dates. They are cheap to install in week three and almost impossible to install in month twenty.
Install in the first ninety days: names and roles for four functions (commercial lead, retail-facing account management, supply and forecasting, marketing execution), and one recurring meeting per function — not one meeting with everybody. Coverage of the middle of their organisation is the asset. Access to the top is not.
Default three: the first order becomes the reference point for every order after it
The first purchase order is treated as a commercial event. It is actually the moment the scale of the relationship gets set, for both sides, in a way that is very hard to move later.
Two things happen at once.
First, the number becomes the baseline against which all later numbers are read. A second order at the same size looks like stagnation; at half the size, like collapse — even if the first number was simply wrong. Nobody re-derives the original assumption. They compare to it.
Second, and more consequentially: an oversized first order converts a sales problem into an inventory problem, and inventory problems are invisible to you. If the partner takes three quarters of stock on an assumption that does not hold, the months that follow produce no orders, no urgency, and no information. You will read the silence as a market that is slow. It is a warehouse that is full.
The defence is not a smaller first order. It is a recorded first order: the assumption written down next to the number, with a date on which both sides agreed to look at it again. Sell-through at month seven is the natural checkpoint, for reasons set out in The Month-Seven Test — three months of safety stock plus a month of shipping and clearance plus one quarter of actual selling.
Worked illustration, with invented numbers: a partner proposes an opening order covering nine months of their own forecast; you agree to six, and write down that the six assumes two chain listings in the first quarter. If one listing lands, nobody argues about whether the market is bad. The assumption is on the page, and only one of its two legs failed.
Install in the first ninety days: the opening quantity, the two or three assumptions it rests on, and the date the assumptions get read back. One paragraph. It costs nothing, and it is the only thing that makes month seven a diagnosis instead of an argument.
Default four: the order in which retailers hear about you
Nobody writes this one down, and it is close to irreversible.
The first three or four accounts a partner opens set your price position in that market, because retailers talk, and because the terms given to the first account become the floor for the next. A partner under pressure to show early volume will open the easiest doors first — which are usually the most price-aggressive doors, and often the ones with the least brand-building capability.
The consequence shows up a year later as a pricing complaint, not a sequencing complaint, which is why the cause is almost never traced. It is the same mis-attribution described in Your Marketplace Store Is Four Rights, Not One Decision: the problem presents as price, and the decision that created it was about something else entirely.
The corridor argument from Stop Pricing One Export Market at a Time applies inside a single market too. What you are protecting is not a price. It is the distance between the first visible price and every price after it.
Install in the first ninety days: a named list of the first five accounts, in order, agreed before any of them is approached — plus a short list of accounts that require a conversation first. This is not a restriction on their commercial freedom. It is the only version of channel strategy that exists before there are channels.
Default five: what you do at day sixty, when nothing has happened yet
Day sixty is the quietest point in the whole relationship, and the most informative.
The contract is signed, the stock has shipped or is shipping, registrations are pending, and there is no sales data yet. There is nothing to review. Most brands use the gap to work on something else, and schedule the next real conversation for the end of the quarter.
That gap is where the behavioural signals arrive — and behavioural signals always arrive before numerical ones, as argued in Distributor Performance Evaluation. At day sixty you can already observe, at no cost:
- whether the people you were introduced to are still the people answering
- whether questions get answered in two days or eight
- whether bad news arrives before you ask, or only after
- whether anyone on their side has asked you a question about next year
- whether the retail conversations they described as "in progress" have names attached yet
None of this requires data, permission, or a formal review. It requires a note, and someone to write it down. Almost nobody does, which is why the first genuine warning in most partnerships is a missed forecast nine months later, by which time three quarters of the useful evidence has been collected and discarded.
Install in the first ninety days: a day-sixty call with no agenda item called "performance", and a one-page note of the five observations above. Date it. Its only purpose is to be re-read at month seven.
Why none of this can be renegotiated later, and what that means for how you ask
Everything in the five defaults above is cheap now and expensive later, for one shared reason: it concerns the things the contract does not specify.
Hart and Moore's point about ownership is that contracts are necessarily incomplete, so what matters is who holds residual control — the right to decide in the situations the contract never described. Reporting format, who attends which meeting, which account gets opened first, what happens on a quiet day sixty: none of these appear in a distribution agreement. They are settled by practice. And practice, once established, belongs to whoever established it.
So the first ninety days are not a period in which you enforce the contract. They are the period in which practice is created, and practice is where the residual control actually sits.
There is a second reason, and it changes the tone of the asking. Kumar, Scheer and Steenkamp found that in reseller relationships, procedural fairness — whether the process by which decisions get made is seen as fair — predicts relationship quality more strongly than distributive fairness, the share of the value each side receives. Partners tolerate a worse deal arrived at through a transparent process better than a better deal handed down.
That is the practical instruction. Install the defaults as process, in the open, with the reason stated — not as conditions discovered after signature. A reporting format introduced in week two with an explanation is process. The same format introduced in month eighteen is an audit, and will be resisted for reasons that have nothing to do with the format.
The five-line version
Your turn
Sources
- Benjamin Klein, Robert G. Crawford & Armen A. Alchian, "Vertical Integration, Appropriable Rents, and the Competitive Contracting Process", Journal of Law and Economics 21(2), 1978, pp. 297–326.
- Oliver E. Williamson, "Transaction-Cost Economics: The Governance of Contractual Relations", Journal of Law and Economics 22(2), 1979, pp. 233–261.
- Jan Johanson & Jan-Erik Vahlne, "The Uppsala Internationalization Process Model Revisited: From Liability of Foreignness to Liability of Outsidership", Journal of International Business Studies 40, 2009, pp. 1411–1431.
- Jeffrey H. Dyer & Harbir Singh, "The Relational View", Academy of Management Review 23(4), 1998, pp. 660–679.
- Oliver Hart & John Moore, "Property Rights and the Nature of the Firm", Journal of Political Economy 98(6), 1990, pp. 1119–1158.
- Nirmalya Kumar, Lisa K. Scheer & Jan-Benedict E. M. Steenkamp, "The Effects of Supplier Fairness on Vulnerable Resellers", Journal of Marketing Research 32(1), 1995, pp. 54–65.
All figures in this article are illustrative. Countries, companies, brands, individuals and proprietary figures have been removed; only the structure remains. This is not legal advice.
Written from the brand side of this table — twenty-two years appointing and managing distribution partners across Asian markets, across categories and price tiers.
Previously in this series: How to Evaluate a Distributor Before You Sign · Exclusive or Not · When to Stop Selling Cross-Border and Set Up Locally · What to Do When Your Distributor Asks for a Lower Price · When Is a Market Ready for a Local Launch? · Marketing Money in a Distribution Agreement · When Your Distributor Stops Paying · The Month-Seven Test · How to Manage the Distributors You Never Visit · Parallel Imports Are a Price Gap, Not a Piracy Problem · Stop Pricing One Export Market at a Time · Your Marketplace Store Is Four Rights, Not One Decision · Distributor Performance Evaluation