# A one-page account plan for your top five customers, kept in Pipedrive instead of a deck

https://salesbond.app/blog/key-account-plan-small-sales-team/
Published: 2026-10-02 · Updated: 2026-10-26

A key account plan is one page per key customer: what they bought, who on their side is involved, where the account could grow, what could break the next renewal or order, and the next action with a date and an owner. A team of three to ten reps needs about five of these, one for each customer the business cannot afford to lose, and a twenty-minute review of each plan once a quarter. Everything on the page already exists in Pipedrive as won deals, people, notes and activities. The plan is the act of reading it in one place and writing down what you intend to do about it.

That is the whole method. The rest of this article is how we keep the page inside the organization record, so it does not turn into a slide deck that gets opened once in January.

## Why your top five customers need a plan, not a pipeline stage

Open the pipeline of any small team and look for the customer who pays the most. Usually they are not there. Their deal was won a year ago and lives in the closed section, and the only thing with their name on it is an organization record nobody has a reason to open until something goes wrong.

That is not a CRM failure. A pipeline tracks a decision in progress, and a key customer is not in the middle of a decision for most of the year. They are using what they bought. So the tool that is built around the deal goes quiet on exactly the accounts that matter most, and the rep who sold them is busy with the deals that are moving. The mirror task, [a named target account list in Pipedrive](https://salesbond.app/blog/target-account-selling-pipedrive/) for companies that are not customers yet, is a different job.

The enterprise answer to this is a strategic account management program. Gong’s guide to it walks through eight steps, from identifying strategic accounts and assigning dedicated account managers to building a full profile of each one and creating executive alignment ( [Gong: a complete guide to strategic account management](https://www.gong.io/blog/strategic-account-management)). Every step is sensible. Every step also assumes a person whose job is accounts. On a team of five where each rep both hunts and farms (we wrote about [whether a small team needs separate hunters and farmers](https://salesbond.app/blog/hunter-vs-farmer-sales-small-team/)), that person does not exist. So we shrink the artifact instead of the program: one page, written by whoever owns the account, read by the manager once a quarter.

A page does something a stage cannot. A stage says where a deal is. A page says what we believe about a customer: that their second site will need licenses next year, that the champion is interviewing elsewhere, that nobody on our side has ever spoken to the person who approves the budget. Those beliefs are what the next expansion deal or renewal is built from, and until that deal exists they have nowhere to live in a pipeline. Which of those beliefs is strong enough to act on is a separate question: [what should trigger an upsell or cross-sell conversation](https://salesbond.app/blog/upsell-vs-cross-sell-pipedrive/).

## One-page key account plan: who, what they bought, what could grow, what could break

We tried the long version first. A template with a SWOT, a stakeholder map, a relationship score out of ten and an “executive summary” for a customer with eleven employees. It was filled in once, during planning week, and never touched again. The page that survives has five blocks, and each one is short enough to be rewritten in a review.

1. **Who.** The people, by name and role: who signed, who uses it every day, who approves money. And the gap, written as a name or a job title: the person we have never spoken to.
2. **What they bought.** The shape of the purchases rather than a copy of the invoices: what, when, how often. For a customer on an annual contract this is one line. For one who reorders it is a rhythm, and the rhythm is the useful part.
3. **What could grow.** One hypothesis, phrased as a sentence the rep could say on a call. “The Leeds office is on a spreadsheet and the ops lead has complained about it twice” is a hypothesis. “Upsell potential: high” is not.
4. **What could break.** One or two risks in plain words. The champion is changing jobs. A competitor is already in another department. The last price conversation went badly and nobody followed up.
5. **Next action, with a date and an owner.** The only line that must never be empty.

Nothing else. No relationship strength, no account tiering formula, no diagram. A number that did not come out of the CRM is an opinion in disguise, and the manager will read a plain sentence faster than they will interpret a seven. If you run Salesforce, [Salesforce Account Plans](https://salesbond.app/blog/salesforce-account-plans-manager-check/) give that page a native home, and we wrote down what a manager should check before the team starts using them.

![Key account plan laid out as five blocks: who, what they bought, what could grow, what could break, next action with a date and an owner](https://salesbond.app/wp-content/uploads/2026/10/key-account-plan-small-sales-team-2.png)

## Pull it from Pipedrive in ten minutes

The page is the organization record. We do not keep it anywhere else, because the organization is the one object in Pipedrive that outlives the deal.

Start with the fields. Pipedrive lets you add custom fields to deals and leads, people, organizations, products and projects; creating or editing them needs the right global permission, and how many you can have depends on your plan ( [Pipedrive: custom fields](https://support.pipedrive.com/en/article/custom-fields)). We add two text fields to the organization, **Growth hypothesis** and **Risk**, and we leave it at two. This Risk field is a belief about the whole account, not the risk reason we put on a renewal deal. The organization detail view shows the address, website, any custom field you added and the contact label ( [Pipedrive: detail view](https://support.pipedrive.com/en/article/detail-view)). The “who” block does not need a field either: the people are person records linked to the organization. The missing person goes in as a person record with a job title, even if all you have is the title.

Mark the five. Pipedrive lets you put a contact label on an organization, and once you add the label column to the organization list, that column can be sorted and filtered, so a “Key account” label gives you the five on one screen ( [Pipedrive: contact labels](https://support.pipedrive.com/en/article/contact-labels)). One catch: an organization carries one label (multiple labels are in beta for a closed group of users), so “Key account” will replace whatever label it had before, such as “Customer”. Add the two custom fields as columns in that list and you are looking at every plan at once, which is as close to a key-account report as a small team needs.

Then the purchases. Every won deal with this customer was made against their organization, so “what they bought” is a minute of reading their deal history rather than a field to maintain. Where a customer has several sites or subsidiaries buying separately, link them with Related organizations from the detail view, choosing Parent, Daughter or Related, so the plan covers the whole customer and not the one branch that happened to sign first ( [Pipedrive: related organizations](https://support.pipedrive.com/en/article/related-organizations)).

The same Parent link is what you use [when the buyer is acquired mid-deal](https://salesbond.app/blog/prospect-acquired-mid-deal-pipedrive/), instead of merging the two companies.

The next action is the part teams get wrong. It has to be an activity linked to the organization, not to a deal, because for most of the year there is no open deal to hang it on. Pipedrive lets you link an activity to a person, organization, lead, deal or project, from the activity tab of the organization’s detail view ( [Pipedrive: activities](https://support.pipedrive.com/en/article/activities)). A call or a task with a date. When the date passes it shows up in the rep’s activity list like any other overdue item, which is the whole point: the account plan gets the same nag the pipeline gets.

If you are on Growth or higher, one more thing helps the “who” block. The contacts timeline shows every organization with its open deals, emails, notes and completed activities on one page, and you can [set a follow-up frequency](https://salesbond.app/blog/pipedrive-contacts-timeline-follow-up-frequency/) so that any contact not touched within that window turns red ( [Pipedrive: contacts timeline](https://support.pipedrive.com/en/article/contacts-timeline)). Set it to the gap you would be embarrassed by, and the five key accounts tell you themselves when one has gone quiet.

Ten minutes per customer, the first time. After that the fields change only when the belief changes.

## The quarterly 20-minute account review

The account owner and the manager, one organization open on the screen, twenty minutes, and five questions that match the five blocks. We do it as its own slot, not inside the team QBR; [the small-team QBR](https://salesbond.app/blog/sales-qbr-small-team/) is about the team’s numbers, and a key customer deserves not to be the last item on that agenda.

The questions are blunt on purpose.

Who did we actually talk to this quarter, and is the gap still a gap? If the “person we have never spoken to” is the same name as last quarter, that is the the first action for the next one, and the rest of the review can wait.

Did they buy anything, and did the rhythm hold? For a reorder customer, the owner should be able to say when the last order was and when the next one is expected, from the deals list, without opening a spreadsheet. To see across the whole book [which of your customers have gone quiet](https://salesbond.app/blog/pipedrive-inactive-customers-accounts-filter/), one account filter and one date per customer are enough.

Is the growth hypothesis still the same sentence? A hypothesis that has not moved in two reviews is decoration. Either delete it or turn it into a deal. This is the direction that matters: the plan feeds the pipeline. When the ops lead finally asks for a quote for Leeds, a deal gets created from the organization, with the hypothesis as its first note, and the pipeline takes over. The plan never tries to be the pipeline.

Has anything new appeared on the risk line? The manager’s job here is to ask what the rep has heard, not what the rep has written. A champion’s resignation usually reaches the CRM a month after it reaches the rep’s inbox.

What is the date on the next activity? If there is none, the review is not finished.

For a customer on an annual contract, the renewal itself does not belong on this page; it belongs as a deal in [a renewal pipeline with the contract end date on it](https://salesbond.app/blog/customer-renewal-process-pipedrive/). The account plan is the reason you win that renewal. The renewal pipeline is the mechanism that stops you forgetting it exists.

For [customers who reorder](https://salesbond.app/blog/repeat-orders-pipedrive-new-deal-per-order/) rather than sign, the review has one more input: whether the next order is on time. A disclosure, since this is where our product fits: we build SalesBond, a sales management layer on top of the CRM a team already uses; we started with Pipedrive. Its Customer Health works out each repeat customer’s buying cycle from their won deals and shows whether they are on track, overdue or leaving, so you can put the overdue ones first on the review list instead of discovering them during it. [Check your CRM with SalesBond](https://salesbond.app/) if you want that list without building the filter yourself. What it cannot tell you is why an account is late; that is still the twenty minutes.

## When an account plan is overkill

There are customers who should not have one, and a team that writes plans for everyone has written none.

Skip it for one-off sales. If a customer buys once and there is no second purchase to plan for, “what could grow” is fiction and “what could break” is already over. The organization record with a won deal on it is the whole story.

Skip it when you cannot name the five. If revenue is spread evenly across forty customers and no five of them stand out, the problem is not account planning. It is deciding who owns which accounts at all, which is [territory planning for a small team](https://salesbond.app/blog/sales-territory-planning-small-team/), and the plan can come later for whichever accounts turn out to be the ones you would lose sleep over.

Skip the written version, too, if it has quietly become a report for the manager rather than a tool for the rep. The test is simple: does the rep open the organization record before calling the customer? If the answer is no and the fields are being filled in the night before the review, the page has become homework. Cut it back to teh next action and the gap, and rebuild from there.

And do not let five become fifteen. The number is not sacred but the moment the list is longer than the owner can hold in their head, it is no longer a list of key accounts. It is the customer base, and the customer base is managed by rhythm and dates, not by a page each. Five pages, twenty minutes a quarter, one activity with a date on every one of them. We have not found a version of account management that is smaller than that and still works, and for a team of five we have not needed one that is bigger.