Ask a founder why they lost their last big deal and you get an answer in about two seconds. Price. It is almost always price. Then you call the buyer who actually made the decision, and price comes up fourth, behind a proposal that took too long, a website that did not prove the company had done this work before, and an internal champion who never had anything to forward to their boss.
That gap is the whole reason to run win-loss analysis. Your last twenty deals already contain the answer to what is wrong with your marketing, your pricing, and your sales process. Nobody wrote it down.
Why the CRM reason codes are lying to you
Every CRM has a closed-lost reason field. HubSpot has one. Pipedrive has one. Salesforce has one with fifteen options. Look at yours right now and you will find that price is the most-picked answer, probably by a wide margin.
Here is the problem. The salesperson fills in that field, not the buyer. And of all the available answers, price is the only one that does not reflect badly on the person filling it in. "We lost on price" means the market is hard. "They could not tell what we do" means the salesperson did not explain it. "Our proposal was late" means someone dropped it. The dropdown is a record of what your team believed, and that is a different thing from what happened.
I am not saying your salespeople are dishonest. They are guessing, under time pressure, about a conversation that happened inside a company they cannot see into. The buyer's real decision meeting happened without them. Everybody in B2B is guessing at that meeting. Win-loss analysis is just the decision to stop guessing and ask.
Who to call, and when
Ten to twelve interviews per quarter is enough for a business under 50 people. You are not running a statistical study. You are listening for the same sentence to come out of three unrelated buyers who have never met each other. That repetition is the signal, and it usually appears by the eighth call.
Split them roughly like this: six deals you lost, four you won, two that went dark and never decided anything. The no-decision group is the one everyone skips and it is often the most useful, because "we did nothing" is a real competitor and usually your biggest one.
Timing matters more than people expect. Call two to four weeks after the decision. Any earlier and the buyer may still be inside procurement, or feel awkward telling you no twice. Any later than about six weeks and the details compress into a tidy story that is easier to tell than the messy truth. In that middle window you still get specifics: which page they read, who pushed back internally, what your competitor said in the final meeting.
Will they take the call? More often than you think. Around a third to a half say yes when the ask is short and honest. The framing that works is plain: we are not trying to win the deal back, we want to get better, and it will take twenty minutes. Do not offer a gift card. It changes the tone and it attracts people who want a gift card.
Who should make the call
Anyone except the salesperson who worked the deal.
Buyers will not tell the person they rejected the real reason. They are being polite. They are avoiding an argument with someone who might try to reopen the deal. So they say price, the person writes down price, and the loop closes on itself with nobody any wiser.
In a company under 50 people, the founder making the call personally gets the best response rate and the most candid answers. Buyers will help an owner who asks directly in a way they will not help a research firm. If the founder cannot do it, a marketing lead or an operations person works fine. The only requirement is that the interviewer had no stake in the outcome.
The eight questions
Order matters as much as content. Start at the beginning of their process, not at the decision, because the moment someone begins defending a choice the answers turn rehearsed.
- What was going on in the business that made you start looking for this in the first place?
- How did you find the companies you looked at?
- Who else was involved in the decision, and what did each of them care about?
- What did you look at on our website or in our materials before we spoke?
- Walk me through the moment you narrowed it down. What was still open at that point?
- What did the company you chose do or say that we did not?
- Was there a point where you almost went the other way?
- If you were advising us, what is the one thing you would change?
Question four catches most of the marketing problems. Buyers will describe pages that confused them, proof they went looking for and could not find, and industries they wanted to see named. Question seven is where the honest answer usually lives, because it invites the buyer to describe a moment of doubt rather than deliver a verdict.
Record the call with permission. Do not take notes while listening, because you will steer the conversation toward what you are writing down.
Turning answers into changes
A stack of interview notes is not a result. The exercise only pays off when each finding gets routed to whoever owns it.
Messaging problems go to the website and the sales deck. If four buyers say they could not tell whether you had worked in their industry, that is a case-study problem, and our guide to writing B2B case studies covers the format that answers it. Positioning problems go into the messaging framework; win-loss is the input that keeps positioning work honest, because a workshop tells you what the team believes while interviews tell you what buyers actually said. When the two disagree, the buyers are right.
Process problems are usually the fastest to fix and the most embarrassing to hear. Proposals taking nine days when a competitor turns them around in two is a calendar problem, and it can be fixed this month.
Product and capability gaps get one of two responses: put it on the roadmap, or turn it into a disqualification rule. That second option is underrated. If you keep losing the same shape of deal to the same competitor over a capability you have no plans to build, stop competing for those deals and say so in the first call. Losing in week one costs almost nothing. Losing the same deal in week eight, after a proposal and three meetings, is expensive.
The quarterly rhythm
Run this every quarter, not once. A one-time win-loss project produces a document. A quarterly habit produces a feedback loop, and the loop is the point.
The output should be one page. Top three patterns, the evidence for each, the specific change being made, and who owns it. Show that page at the quarterly review beside the pipeline numbers. When a finding from last quarter turns into a change that shows up in this quarter's marketing dashboard, people start taking the interviews seriously and the calls get easier to schedule.
Total cost is about six to eight hours a quarter, done in-house. Outside firms in Canada charge roughly CAD $1,000 to $2,500 per interview for a fully managed program, which rarely pencils out at SMB deal sizes. For a business under 50 people, do it yourself, and have the founder make the calls.
The takeaway
You are already running the experiment. Every deal that closes is a data point about your positioning, your pricing, and your process, and right now that data is being thrown away and replaced with a dropdown. Twelve calls a quarter recovers it.
Call the buyers. Ask what happened. Write down what they say, not what you hoped they would say. It is the cheapest research you will ever run, and it is the only kind that comes from the people who actually decided.