Customer discovery interviews, and how many are enough
B2B customer discovery interviews: the sample size the research converges on, why your segments change that number, and the questions that get honest answers.
Read
Deals rarely die from a bad proposal. They die because nobody wrote down what had to happen between the good meeting and the signature. One shared document fixes that, and tells you early which deals were never real.
Most deals do not die from a bad proposal. They die because nobody ever wrote down what has to happen between the good meeting and the signature, so the steps that needed four weeks were discovered in week six. A mutual action plan is the unglamorous fix: one shared document, built with the buyer rather than for them, listing every remaining step, who owns it, and when it is due.
The reason to use one is not that it closes deals faster. It is that it tells you early which deals were never real. A buyer who will not agree a date for a security review is giving you information, and it is cheaper to have that conversation in week two than at the end of the quarter.
A mutual action plan is a close plan the buyer has seen and edited. That second half is the whole thing. A close plan you keep in your CRM and never show anyone is a forecast. The same plan opened in a call, with the buyer adding the two steps you did not know about, is a mutual action plan.
It is also not a project plan. It covers the path to a decision, not the path to a working deployment. The implementation plan comes after, it involves different people, and merging the two produces a document nobody reads.
Outreach publishes a template in its guide to mutual action plans with six columns, and the set is a reasonable floor: objective, milestones, owner, deliverables, status, success metrics. The structure is not controversial. What matters is which column you refuse to leave blank.
It is owner, and specifically owner on the buyer side. A milestone with a date and no named buyer attached to it is yours alone, which means it is not mutual and will slip without anyone feeling responsible. The useful question in the call is not when can we do the security review, it is who on your side owns the security review. If the answer is vague, that is the finding.
| Field | What it is for | How it fails |
|---|---|---|
| Objective | The buyer's business goal in their words | Written in your product language, so the buyer never adopts it |
| Milestones | Every remaining step to a decision | Only your steps, so procurement and legal appear late |
| Owner | A named person per step, both sides | Left blank on the buyer side, which makes it a forecast |
| Deliverables | What has to exist for the step to be done | Vague, so the step is argued about rather than finished |
| Status | Current state, visible to both | Updated by you only, so it records your optimism |
| Success metrics | How the buyer will judge it worked | Omitted, so the business case cannot be defended internally |
The common failure is building a careful plan with one enthusiastic champion and treating that as coverage. It is not. Demandbase's research on buying groups and revenue puts a typical buying group at 13 to 17 stakeholders, which is more people than any single champion can brief accurately on your behalf.
The same research reports something more uncomfortable about spreading yourself: win rates peaked at around 29% when teams concentrated on three buying groups and fell to roughly 12% at six. Read carefully, that is an argument for fewer accounts with real coverage rather than more accounts with a named contact each. It is also vendor research, which matters, and we come back to that below.
Practically, the plan should name the people who can stop it. Who signs, who reviews the contract, who owns the budget line, who has to integrate it. If your plan has three rows and your buying group has fourteen people, the plan is describing your conversation rather than their decision. Our note on the B2B buying committee and who is really in the room is the companion piece to this one, and worth reading first if the names are not obvious.
You will see a figure saying mutual action plans lift win rates by 26%. It comes from the Outreach guide linked above. It is plausible and it is also not evidence you should lean on, for one specific reason: the article states the number without a sample size, a time period, or a methodology.
There is a deeper problem that no vendor figure in this space solves. Reps who run mutual action plans are, on average, more disciplined reps working better qualified deals. Any comparison between deals with a plan and deals without one is partly measuring that, not the document. Nobody has published a study that separates the two, and until somebody does, the honest claim is that a mutual action plan makes a deal legible, not that it makes it more likely to close.
Legibility is worth plenty on its own. It is the difference between a forecast built from dates the buyer agreed and one built from how the last call felt, which is the distinction our piece on forecasting B2B sales and starting with slippage works through.
The mechanics that decide whether it survives contact:
The questions that surface the hidden steps are mostly discovery questions asked late, and the eight in our guide to sales discovery call questions work as well in week six as in week one.
It is overhead, and overhead is not always paid back. On a small transactional deal with one decision maker and a two week cycle, a plan with six columns is theatre that slows you down. The rough test is whether the deal has more than one gate you cannot see from here. If it has one buyer, one signature and no procurement, skip it.
It also does not work with a buyer who will not co-own it. If you have asked twice for owners on their side and received nothing, you do not have a plan problem. You have a deal that is not yet real, and the correct response is to stop building documents and go find the person who can commit. Pushing the plan harder in that situation is the most common way it becomes a chase list.
Where it earns its keep is the long deal with multiple gates, which is most enterprise work and nearly all of it in the Gulf. Once procurement is involved the plan becomes the only place the real timeline lives, and the levers that move those dates are set out in negotiating with procurement departments.
Both sources behind the numbers in this piece are vendors selling into the problem they are measuring. The Demandbase buying group benchmarks are described as roughly one year of data, framed by Demandbase as correlations rather than causes, with no sample size given. The Outreach win rate figure has no stated methodology at all. We could not find peer reviewed work isolating the effect of a mutual action plan on deal outcomes, and we looked.
So treat the figures as directional and the mechanism as the point. A document that forces a named owner and an agreed date onto every remaining step will tell you things about a deal that no amount of pipeline review can. That is true whether or not the lift is 26%, and it is the reason to write one.
Send us three open deals and, for each, the date the buyer named and who on their side owns the next step. If you cannot fill in the second column for any of them, we will tell you which deal to stop forecasting and what to ask on the next call.
Book a Free Consultation