100 Accounts Each. 30 Shared Accounts. None Were Top of Mind.

Two partner companies ran an account mapping exercise. Each side uploaded roughly 100 accounts. OnlyCommon identified 30 accounts present on both lists, and the comparison took around two minutes.

Two minutes is the least interesting fact in that paragraph. Speed only matters when the slower method produces the same answer, and here it did not: none of those 30 shared accounts had been top of mind for either team during their normal account-mapping conversations. The partners had talked about their customers and prospects. They knew each other well enough to be discussing a partnership seriously. The overlap that emerged still consisted of accounts neither side had thought to raise.

This is a single anonymized example, so it is not a benchmark and it says nothing about what a typical overlap looks like. The verified figures behind it are set out in the customer story. What it does illustrate is a mechanism that most partnership teams will recognise once it is named.

What memory-based mapping is good at, and what it is not

The standard way two companies start looking for shared accounts is a conversation. Two partnership managers get on a call, establish that the fit looks plausible, and try to make it concrete by trading account names. Someone takes notes, a handful of recognisable logos come up, and both sides go away to check internally.

That conversation is valuable, and it is worth keeping. It surfaces relationship history, competitive context, executive sponsorship, past attempts, timing constraints — the kind of detail that no dataset contains and no matching engine can infer. It is also how two teams build enough trust to do anything together at all.

What it is not good at is completeness. Ask anyone in sales or partnerships to name accounts from memory and you get a predictable subset of the account base: the largest customers, the deals closed most recently, the logos that appear in board decks, the deals discussed in this week's pipeline review. What tends not to surface is the long middle of the account list — steady mid-market customers renewing without drama, accounts inherited from a rep who moved on, prospects that went quiet a year ago, regional accounts only one territory thinks about. In many companies that middle is most of the list.

So when two partners compare from memory, they are comparing two partial views. The accounts each side recalls are real, but they are a sample, and the intersection of two samples is smaller and less predictable than the intersection of the two full datasets. That is the difference the 30 accounts in this example expose.

The precise distinction

It is worth stating the difference in one line, because it is easy to blur:

• A partnership conversation asks: which accounts can we think of that we might have in common?
• Systematic matching asks: which accounts do our two datasets actually share?

Both questions are legitimate. Only the second one has an answer you can verify, and only the second one produces a list you can be confident is not missing most of itself. How partner account mapping actually works walks through the mechanics of asking it properly.

The practical consequence is that a memory-based exercise can never tell you how much you missed. Two partners who recalled three shared accounts and two partners with thirty shared accounts look identical from the inside if recall was the only method used. Completeness is not something a conversation can establish about itself.

Thirty shared accounts is not thirty opportunities

This is where account mapping is most often oversold, so it is worth being blunt. The 30 accounts in this example were 30 overlaps to investigate. They were not 30 co-sell opportunities, and describing them that way would misrepresent what a match means.

A shared account means both companies have a record or a relationship tied to the same organisation. It says nothing about whether either relationship is strong, whether the account is a paying customer or a stalled prospect from two years ago, whether anything is happening there now, or whether a joint conversation would help the buyer rather than confuse them. Those are human judgements, and they have to be made account by account, usually by the people who own them rather than by the partnership team alone.

It is also worth being honest about matching itself. Company data is messy: names differ, subsidiaries and legal entities complicate things, domains do not always resolve the way you expect. No matching process is perfect, and a shared-account list is a strong starting point rather than an authoritative census.

A sequence that works

The reason the discovery step matters so much is that it is the only step that cannot be repaired later. You can revisit a prioritisation call; you cannot act on an account nobody knows is shared. In practice a workable sequence looks like this.

Start by discovering the full overlap from the data rather than from recall, so the list you are working from is as complete as your two datasets allow. Then review those accounts with the people who hold the relationships on both sides, adding the context the data does not carry: who the customer is, what state the relationship is in, what has already been tried. Prioritise from there — a small set where both sides have something real and where a joint conversation is defensible from the buyer's point of view. Name an owner on each side for each of those accounts, so there is a person rather than a shared responsibility. Only then coordinate the introduction, the joint call, or the co-sell motion.

The common failure is inverting the first two steps: prioritising a list before finishing the discovery, which means prioritising within a sample. The second most common failure is trying to act on the whole overlap at once. Thirty accounts split across two companies with no prioritisation tends to become thirty half-conversations. Five well-chosen accounts produce something you can learn from, and evidence from a small pilot is what earns budget for co-selling programmes and integrations later.

If you want a framework for deciding whether the overlap you find justifies further investment, how much account overlap makes a partnership worthwhile is a more careful treatment than any percentage rule of thumb.

Why partners rarely just exchange spreadsheets

There is an obvious route to completeness: both sides export everything and email it across. It solves discovery and introduces a different problem.

Account lists are commercially sensitive, and many companies are reluctant to hand a full export to another organisation — particularly one that may overlap competitively now or later. Even where nothing prevents it, the exchange invites a longer internal conversation than the exercise itself warrants.

It is also unnecessary. The question is where two account sets intersect, and answering it does not require either side to see the other's non-matching rows. That is the shape OnlyCommon works from: each partner uploads their own list independently, and the result shows the accounts the two lists have in common while rows that do not match are not revealed to the other partner. It is not a CRM and it does not replace the systems where your account data lives; it compares two lists and shows the intersection.

That difference is usually what makes the exercise straightforward to get approved. Comparing lists and seeing only what you share is a much shorter internal discussion than sending a customer base to a partner. Teams who want to run it without a systems project can start from exported data — see account mapping without CRM integration.

What this example supports, and what it does not

It supports one claim: two partners who had already discussed their customers found 30 shared accounts that had not been top of mind, in about two minutes. That is a statement about the limits of recall as a discovery method, and it is the only conclusion a single case can carry.

It does not suggest that a 30% overlap is typical, that 30 shared accounts is a normal result, or that any of those accounts produced pipeline or revenue. Overlap varies enormously with market, segment and go-to-market model. Another pair of partners might find three shared accounts, or two hundred. What generalises is the method, not the number.

The argument for running the comparison is not the expectation of a particular result. It is that until the two datasets are compared, neither side knows which situation they are in — and every decision made in the meantime rests on what two people happened to remember.

If you are at that stage with a partner, comparing the lists is a short exercise: each side uploads independently, only the shared accounts are revealed, and the conversation that follows starts from evidence instead of recall. You can map a partner free, or read more about how partner account mapping software fits alongside the systems you already run.

Frequently Asked Questions

How many shared accounts is a good result in partner account mapping?

There is no universal percentage. A 30% overlap between two focused lists can be far more useful than a 5% overlap between two large databases, and the reverse is equally true. What matters is whether the shared accounts sit in segments you both sell to, whether either side has an active opportunity or a strong relationship, and whether your teams can realistically act on them this quarter. Judge the overlap by what it makes possible, not by the size of the number. How much account overlap makes a partnership worthwhile works through that assessment in detail.

Does finding a shared account mean there is a co-sell opportunity?

No. A shared account means both companies have a record or relationship tied to the same organisation. That is a signal worth investigating, not a qualified opportunity. Before anything happens commercially, someone has to check how strong each side's relationship actually is, whether the account is a customer, a stalled prospect or a closed-lost, whether the timing makes sense, who owns it internally, and whether a joint conversation would genuinely help the buyer. Some shared accounts become co-sell motions. Others are best left alone.

Why do memory-based partner conversations miss shared accounts?

They are not designed for completeness. When two partnership managers trade account names on a call, they answer with what they can recall: recent deals, large customers, strategic logos, whatever came up in the last pipeline review. Those conversations are genuinely useful for context and relationship detail, but recall covers only part of an account base. Mid-market customers, inherited accounts and quiet renewals rarely come to mind, so shared accounts sitting in that part of the list can stay undiscovered until the two datasets are compared directly.

Do both partners need to share their full customer lists?

No. The question account mapping answers is where two account sets intersect, and answering it does not require either side to see the other's non-matching accounts. With OnlyCommon, each partner uploads their own list independently and the result shows the accounts they have in common; rows that do not match are not revealed to the other partner. Account lists are commercially sensitive and many companies are reluctant to exchange full exports, so that distinction usually makes the exercise easier to approve internally.

How long does partner account mapping take?

It depends on the state of your data and how your teams work, so there is no universal answer. In the anonymized example described in this article, two lists of roughly 100 accounts each produced 30 shared accounts in around two minutes. Preparing an export, agreeing what will be compared, and reviewing the result with the people who own the accounts usually takes longer than the match itself. The comparison is the fast part; the judgement that follows is where the time goes.

Can account mapping work without CRM integration?

Yes. Partners can start with exported account data instead of connecting systems, which means no RevOps project, no security review of a bidirectional sync, and no waiting for both companies to prioritise the same integration. That is often the sensible way to validate a partnership before committing infrastructure to it: find out whether meaningful overlap exists first, then decide whether an integration is justified. Account mapping without CRM integration covers how to run it that way.