Win-loss analysis: how to run it, what to ask and how many interviews

Win-loss analysis is a routine of interviewing buyers soon after they decide, both those who chose you and those who chose someone else or nothing, then comparing the two groups to learn what actually decided deals. It replaces the one-word reason a salesperson types into the CRM with the buyer's own account of the evaluation. MaRS describes it as research into why you win or lose a specific sale, with findings that feed product, pricing, positioning, sales and competitive work.

Running it well comes down to five choices: which deals to study, how many interviews to aim for, who conducts them, what to ask, and how to compare the answers. This page covers each. For the conversation with one lost buyer in depth, see how to interview customers who chose another product.

A worked example

Suppose you sell inspection software to commercial construction contractors and close about 120 deals a quarter: 30 won and 90 lost. All figures here are hypothetical. Product Marketing Alliance gives the two basic formulas:

Win rate = won opportunities ÷ total opportunities = 30 ÷ 120 = 25%
Win/loss ratio = won opportunities ÷ lost opportunities = 30 ÷ 90 = 0.33

These numbers tell you how often you win. The interviews tell you why, and whether the reasons differ by competitor, deal size or segment.

Choose which deals to study

Start narrow. Clozd, a win-loss vendor, argues that themes vary widely across business units and product lines, and that narrowing the focus, for example to deals against a single competitor, lets you learn more from fewer interviews. For the inspection software company, a good first scope might be mid-sized contractors that evaluated you against your most common competitor.

Include both wins and losses. PMA recommends aiming for an even number of each so the results are not skewed, while noting that lost buyers are harder to get on a call and deserve extra effort. Wins show what you should keep doing and saying; without them you cannot tell whether a complaint from lost buyers also came up in deals you won.

Include "no decision" outcomes if they are common. A buyer who chose to keep using spreadsheets tells you something different from one who chose a competitor.

Decide how many interviews to aim for

There is no fixed sample size, and anyone quoting a precise number should explain the scope behind it. What sources say:

PMA says there is no set number and suggests building a few interviews into each interviewer's week so the program keeps running.
Clozd frames the goal as saturation, the point where new interviews stop surfacing new themes. It says meaningful insights can come from as few as 10 interviews in a narrow segment, and that most of its clients settle into 50 to 200 interviews a year, based mainly on budget. That is one vendor's account of its own clients.

A reasonable starting plan for the example: six wins and six losses in the chosen scope this quarter, then decide whether themes are still changing. Expect to invite many more people than you interview. PMA's rule-of-thumb example is to ask 20 people if you want to speak with five. At that rate, reaching six of each would mean inviting about 24 won and 24 lost buyers, which the example pipeline can support.

Timing matters. PMA suggests interviewing within about four weeks of the decision, and for wins, before the customer is deep into setup so onboarding does not color their memory of why they chose you.

Decide who runs the interviews

The salesperson who worked the deal should not run the interview. Buyers soften criticism when talking to the person it is about, and a rep may be tempted to reopen the deal. PMA suggests product marketing, sales enablement, customer success or product management instead, with a caution that product managers can focus too narrowly on features.

MaRS's interview guidance adds: tell the buyer you are not from sales and not selling anything, explain how the feedback will be used and that it will be kept confidential, and aim for the buyer to do roughly 90% of the talking.

When to use a third party. An outside interviewer can get franker answers, and PMA notes that outsourcing gives you objectivity and a set quota of interviews. It costs more and adds distance from the raw conversations. One middle path is to run the first dozen interviews yourself to learn the ground, then decide whether volume or objectivity justifies outside help. PMA lists firms such as Clozd and Primary Intelligence among those that run programs.

What to ask

Keep a consistent core so interviews can be compared, and follow up freely on anything interesting. Group questions around the decision:

The trigger. What was happening that made you start looking? Why then?
The group. Who was involved in the decision, and what did each person care about?
The options. Which products or approaches did you consider, including doing nothing? How did you find them?
The evaluation. What did you need to see to feel confident? What did you test or check?
The deciding moment. When did it become clear which way you would go? What happened just before?
Price. How did pricing compare with the other options, and did it change the outcome?
The sales experience. What was helpful or unhelpful in how each vendor worked with you?
The counterfactual. What would have had to be different for you to choose otherwise?

Keep the call short. PMA suggests staying within 20 minutes so you do not eat too far into the buyer's day. End with an open question such as "Is there anything we did not ask that mattered?" MaRS notes that some of the most useful points come in the final minutes.

Compare wins and losses

After each interview, write down the main decision factors in the buyer's words, tagged consistently (for example "mobile offline mode," "price per seat," "implementation time"). Then compare how many won accounts and lost accounts mention each factor. Count accounts, not mentions. A factor that shows up in most losses and few wins is worth acting on. One that shows up equally in both is probably not what decides deals.

Share a short summary every quarter with sales, product and marketing: the scope, the number of interviews, the top factors in wins and losses, what changed since last quarter and one recommended action per team. Track the win rate for the same scope over the following quarters, keeping in mind that many things besides your changes affect it.

Your next step

Pick one scope, list the last quarter's closed deals within it, and send invitations to won and lost buyers this week. If you want perspective from buyers outside your pipeline, such as construction IT leaders who picked a competing inspection tool without ever talking to you, search for people by describing their role, or start from the information technology professionals in construction directory page. Instant Expert finds people whose work matches your question. You review them, it sends the invitations, and you pay for each call that gets booked.