Win-Loss Analysis Shows Which Buyers Pick You and Why

Win-loss analysis means interviewing recent buyers about how they decided: the ones who signed with you, the ones who went elsewhere and the ones who bought nothing at all. Done well, it's the cheapest evidence a company can get of who picks it, in the buyers' own words.

Most of what ranks in search for running one comes from firms that sell win-loss programs. So it treats the work as a sales tool: read the results for why deals were lost, then fix the pitch and coach the reps.

Hardly anyone reads the results for who the company wins. Line up the wins and you can see which kind of buyer picks you. Then you can check whether that buyer belongs to the market leadership chose, or to one the company drifted into.

Deals that end in no decision belong in the same read, and better sales calls alone won't fix a pile of them.

Here's how to conduct one, in six steps. A made-up example runs through the last three.

1. The wins belong in the sample, because they show who picks you

Start with a batch of recent decisions and keep every outcome in it: wins, losses and no-decisions. Most programs study only the losses, yet the wins are the part of the sample that shows who picks you.

Both guides cited here say to include won deals. Pragmatic Institute sells product-management training, and its win-loss framework page names no author. It says to "start by identifying won and lost customers for the product or service you want to investigate."

The other is a blog post by Forrester analyst Beth Caplow that points to Forrester's paid reports and gives no figures from its own research. Her post on getting buyer feedback for win-loss says the interviews should cover "both won and lost deals."

2. Someone other than the rep should hear the buyer's account

The person who sold the deal is the wrong one to ask about it.

Buyers soften what they say to a salesperson they've worked with. Pragmatic Institute's framework says interviewees "may not want to share negative feedback or jeopardize their relationship with a salesperson."

The CRM can't stand in for the interview either. The loss reasons logged there are the seller's reading of what happened. Caplow puts the problem directly: "Sellers' interpretations often differ from buyers' actual reasons." She also writes that "CRM data lacks depth."

So pick an interviewer who had no stake in the deal. A colleague outside the sales team can do it, and Forrester's post suggests going further: "Consider using a third party for further objectivity."

3. Every buyer should get the same questions about how they decided

The interview is about the buyer's decision. Your product will come up wherever it mattered to them. Four questions get at how they decided:

• What set off the search in the first place?
• Who else had a say?
• What did you compare, including doing nothing?
• What tipped the decision at the end?

Keep the questions open. Caplow's advice is to "keep questions open-ended to allow buyers to respond frankly."

Then ask the same questions in every interview. Pragmatic Institute says interviewers "should use the same set of questions for each of their win/loss interviews to collect consistent answers." That consistency is what lets you sort the answers in the next step.

4. Sorting the wins by buyer shows whether you're winning the market you chose

Group the won deals by who the buyer is: company size, industry, the role that signed and what set off the search. The biggest group is the buyer who picks you most often.

Forrester's post backs cutting results this way: "By reviewing results by segment or competitor, you can identify areas to discern where win rates are higher or lower."

Then hold that group up against the market leadership chose. If they match, the plan and the customer list agree.

If they don't, the company is being chosen by buyers it never set out to win, and the customer list is quietly making the targeting decision. Choosing a target market on purpose starts with seeing that gap.

Here's a made-up win-loss analysis example of what sorting the wins can turn up. Every company, deal and figure in it is invented.

Say a firm sells project-accounting software. Leadership chose architecture firms of 50 to 200 people as its market. The firm reviews its last 24 decisions: 10 wins, eight losses and six no-decisions.

In the interviews, seven of the 10 wins turn out to be engineering consultancies the firm never targeted. They arrived through referrals. They signed because the software handled billing across several offices. Leadership chose architects, and the biggest group of wins is engineers.

5. Losses show who you aren't built for, and no-decisions show who can't tell you apart

Losses point to the buyers you aren't built for. Look for where they cluster: one competitor keeps winning, or one kind of buyer keeps saying the product isn't meant for them.

In the made-up example, the losses cluster by buyer: five of the eight are architecture firms, the very market leadership picked. Their buyers describe the product as built for engineers. The reps' CRM notes on those same deals say "price."

Caplow's post gives the general reason a logged "price" can mislead: reasons "such as 'budget' or 'pricing' don't fully explain solution gaps, sales missteps, or unmet buyer expectations." Her advice is general and says nothing about this invented firm.

One caution on the numbers: the seven wins and five losses come from different groups of deals, so turning them into percentages and comparing them would mislead.

No-decisions need their own short read. A pile of them usually means buyers couldn't tell you apart from the alternatives, which is a positioning problem as much as a sales one. Capiche's guide to why deals stall when prospects see no difference covers what to do about it.

6. Once the pattern holds, the choice runs two ways: follow your buyers or change for the ones you chose

If the pattern holds up under testing, it leaves a choice between two directions.

The first is to aim the marketing at the buyers who already pick you. For the example firm, that means marketing to engineering consultancies, the buyers who signed for billing across several offices. It follows evidence the firm already has, and it gives up, for now, the market leadership picked.

The second is to change what you offer and what you say, so you can win the buyers leadership chose. For the example firm, that means changing the product and the pitch until architecture firms stop seeing it as built for engineers.

Leadership may have picked architects for reasons the interviews can't see, and this path keeps that choice. It means changing the product, the pitch or both before the architects' answers change.

Which path fits depends on what leadership is trying to build, and only leadership can weigh that.

A few dozen interviews are enough to raise the question

The strongest objection is about sample size. Win-loss is usually run as a sales tool, built on a few dozen interviews. Using it to question which market the company is in can look like reading far too much into far too little.

Twenty-four interviews can't settle which market we should be in.

Agreed. Treat what they show as a question to test.

The interviews are still the buyers' own account of how they decided. That puts them ahead of the CRM, given Caplow's point that sellers' readings often differ from buyers' real reasons. Cutting the results by segment is Forrester's own advice. As the same post says, "Win/loss analysis without buyer input is guesswork."

When wins keep arriving from buyers leadership didn't pick, test that pattern before spending more on marketing aimed at the original target. And if your own wins are pointing somewhere leadership didn't plan, we're happy to talk through what that means for your brand.

The buyers who signed last quarter already know why they picked you. Someone just has to ask them.