Win/loss analysis: understanding deals from your sales calls

A lost deal is almost always closed the same way. The rep opens the CRM, picks a reason from a dropdown, often "Price" or "Bad timing", and moves on. Six months later the lost reason report shows price as the top cause, and nobody knows what to do with it.

That report is not wrong, it is incomplete. It measures what the team remembered about a deal, not what the buyer actually said during the calls. And in between lies exactly what would help you win the next one.

Why the CRM lost reason is not enough

The "lost reason" field has three flaws that have nothing to do with the reps' good faith and everything to do with how it gets filled in.

• It is filled in afterwards, from memory. The rep sums up in one word a cycle of several weeks and several people. What remains is the impression of the end, rarely the cause at the start.
• It is filled in by an interested party. Putting a loss down to price or timing blames no one. Putting it down to a rushed discovery or a decision maker never met does. The bias is not dishonest, it is human.
• It keeps none of the material. A dropdown records a category, never the sentence behind it. You cannot check, qualify or compare it afterwards.

"Too expensive" is rarely a cause. Most of the time it is the conclusion of a buyer to whom the value was never shown.

The good news is that the material already exists. It sits in the discovery calls, the demos, the negotiations. The buyer said there what triggered the need, what they were comparing, who had to sign off and what worried them. A serious win/loss analysis starts by rereading that material rather than memory.

What to capture for each deal

For two deals to be comparable, you need to capture the same elements on each. Six are enough, as long as they come from what was said and you keep the sentence that supports them.

1. The moment the need appeared

What made the buyer start looking now? A hire, a departure, an audit, an incident, a new executive. A need triggered by a dated event moves forward; a need "we have had for ages" often keeps waiting.

2. The decision criteria

What the buyer said they would judge solutions on, in their own words. For instance, as an illustration: "It has to be live before our peak season" or "Our IT will not approve anything not hosted in Europe". A criterion voiced and never addressed in the demo is a loss foretold.

3. Who decides

The people named in the decision process, and the ones you actually met. The gap between those two lists alone explains a good share of the deals that stall for no apparent reason.

4. The competitors mentioned

Who else was being looked at, and when the name came up. A competitor named in discovery does not weigh the same as a name that surfaces in the week of the decision. Do not forget the most frequent competitor: the status quo, the spreadsheet, the in-house tool.

5. The objections, and whether they were handled

Each objection raised, the answer given, and above all how the buyer reacted to that answer. A "handled" objection the buyer brings up again two calls later was not handled.

6. The last next step agreed

What was decided at the end of the last conversation, with a date and an owner, or the absence of all that. A deal that ends on "we will get back to you" was often lost right there, well before the official no.

Reading won and lost deals side by side

The most common mistake is to analyse only the losses. Without the wins next to them, you cannot tell whether an element is specific to lost deals or simply everywhere. A price objection also heard in the deals you won does not explain your losses.

What matters here is not one deal, it is recurrence. An element present in one loss is an anecdote. The same element present in most losses and missing from the wins is something to work on.

What the CRM says, and what the calls say

The sentences on the right are examples, but they show the gap. "Price" becomes an internal justification problem, so a return on investment argument to equip. "Competitor" becomes a missing integration. "No response" becomes incomplete qualification.

The win/loss interview with a lost customer

The calls tell the cycle from the outside. A short interview with the buyer after the decision fills in what they leave out. It works better when run by someone other than the deal's rep, a few weeks after the decision, in twenty minutes at most, and with a clear statement that it is not about reopening the sale.

• What led you to look for a solution at that point?
• Which options did you consider, including changing nothing?
• Which criteria did you finally decide on?
• Who took part in the decision, and who had the final say?
• At what point did you feel your choice was made?
• What was missing on our side for you to go further?
• Is there a question we never really answered?
• If you had to make this choice again today, what would you look at first?

Listen without correcting. The goal is not to be right about the deal, it is to understand how the buyer experienced it. Record the interview with their consent and attach it to the same customer as the other conversations, so it is reread alongside them.

A monthly review, not a yearly project

A win/loss analysis run once a year produces a fine document and few changes. A short monthly review produces the opposite. One hour is enough, with the sales lead and two or three reps, on the deals closed that month.

• List the deals won and lost this month, without exception.
• For each, capture the six elements from the conversations, not from memory.
• Look for what comes back in the losses and is missing from the wins.
• Keep one or two buyer sentences per finding, so the discussion rests on facts.
• Decide on a single change for the next month, and name it.
• The following month, start by checking whether that change was applied.

On a small number of deals, be careful: a handful of cases is not a trend. Add up several months before drawing a strong conclusion, and keep going back to the passages rather than to percentages.

What to change next

A win/loss analysis is only useful if it changes something in the way you sell. Three places hold most of the corrections.

• Discovery questions. If the losses lack a dated trigger or explicit criteria, add the questions that bring them out from the first call.
• Qualification. If the losses never reached the decision maker, make meeting them a condition for moving the deal forward, not a hope for the end of the cycle.
• Messaging and demo. If a criterion or a competitor keeps coming back in the losses, build the answer, place it in the demo and check the following month that it is being used.

Keeping this up without rereading everything by hand

Capturing six elements per deal from the calls takes time if it is all done by hand. That is where we can help. Each conversation attached to a customer file is transcribed then analysed: objections, handled objections, competitor mentions, decision criteria, decision process and next steps are picked out as passages, in the buyer's words.

Deals are detected from what was said, opened, won or lost, and stay editable by hand when the team knows better. The assistant can then answer a question such as "which objections come up most in the deals we lost this quarter?" by counting themes from the tagged passages, each claim pointing to the exact passage.

The tool does not replace the review, it prepares its material. The decision to change a discovery question or a qualification step stays with the team, and it is easier to make when everyone can reread what buyers actually said.

A lost deal costs you once. A deal lost for a reason nobody understood costs you every time it happens again.