Win-loss analysis is the practice of finding out why deals closed the way they did, from the people who decided. The failure is almost never the analysis. It is that the input is a seller's account of a loss, recorded by the seller, at the moment they least want to think about it.
Ask a rep why a deal was lost and "price" is available, defensible and requires no further conversation. Ask the buyer and price is rarely the whole answer, because a buyer who genuinely wanted the thing finds budget.
Why the dropdown lies, structurally
It is not dishonesty. Three forces push the answer toward the same few values regardless of what happened.
- The seller was not in the room when the decision was made. Most B2B decisions are finalised in a meeting the vendor does not attend. Whatever the seller records is an inference from the last conversation they had.
- Some reasons are safe and some are not. "Price" and "timing" reflect on nobody. "We did not understand their requirements" reflects on the person filling the field.
- The field is completed during forecast hygiene, weeks after the loss, alongside twenty other records. The incentive is to clear the queue.
The result is a dataset where price dominates, because price is the safe answer that closes the field fastest. Programmes then act on it, discount, and find that win rates do not move.
What actually produces a usable answer
Interviews with buyers, conducted by somebody who did not work the deal. That is the whole method, and everything else is logistics.
| Source | What it tells you | What it costs |
|---|---|---|
| CRM dropdown | What sellers believe, or what is safe to record | Nothing, and it shows |
| Seller debrief | Useful colour, same bias, better detail | Thirty minutes per deal |
| Buyer interview, internal | Closer, but buyers soften it for a familiar face | An hour, plus scheduling |
| Buyer interview, third party | The unflattering version, which is the useful one | The most, and the reason most programmes stop |
The pattern in that table is that accuracy and cost rise together, and the cheapest option is the one that produces confident, wrong conclusions.
How many interviews, and which
Fewer than people expect, chosen more carefully than people expect. Ten to fifteen conversations a quarter surface the recurring themes; the twentieth rarely tells you something the tenth did not.
Interview wins as well as losses. This is the part most programmes skip and it is where the actionable answers live. A won deal tells you what the buyer thought they were buying, which is often not what you thought you were selling, and that gap is worth more than another loss reason.
Prioritise the losses where you were a finalist. A deal lost at first contact tells you about targeting. A deal lost in the final two tells you about the product, the proof or the commercial terms, and those are decisions somebody can act on.
When not to run a programme
Under about twenty closed deals a quarter, do not formalise this. The founder or the head of sales already knows why each one went the way it did, because they were in most of them. A programme adds process to knowledge that already exists.
If nobody owns acting on the output, do not start. Win-loss produces findings that implicate product, pricing and marketing, none of whom commissioned it. Without an owner who can move those, the programme generates a quarterly document that circulates and changes nothing, and that is worse than not knowing, because it feels like diligence.
If the answer would not change anything, skip it. Teams sometimes run win-loss to justify a decision already made. The test is whether you can name, in advance, a finding that would change what you do next quarter. If you cannot, the programme is theatre.
The order to do it in
- Stop trusting the dropdown as evidence. Keep it for volume, not for causes.
- Pick ten deals, weighted toward finalist losses and recent wins.
- Have somebody outside the deal run the interviews. This is the single change that improves the data most.
- Write the findings against a decision, not a template. What should change, and who owns it.
- Re-run the same questions next quarter so the themes are comparable, which is where the value compounds.
Step four is the difference between a programme and a report. A finding nobody owns is the same failure that turns an undocumented sales motion into a hiring problem: the knowledge exists and reaches nobody who can act on it.
Questions people ask
What is win-loss analysis?
Finding out why deals closed the way they did by asking the people who decided. The useful version asks buyers; the common version asks the CRM.
How many interviews are enough?
Ten to fifteen a quarter surfaces the recurring themes. The number matters less than who runs them and which deals you pick.
Should we interview wins too?
Yes, and most programmes do not. A win tells you what the buyer thought they were buying, which is frequently not what you thought you were selling.
Can the rep do the interview?
They can, and the buyer will soften it. Someone outside the deal gets the version that is worth hearing.
Is the CRM closed-lost reason useless?
Not useless, but it measures what sellers record rather than why buyers chose. Treat it as a volume signal and never as a cause.
Or skip the search
Want the person who owns it after it ships?
If the work is a standing obligation rather than a bounded project, tell me which system you want built first. You get example profiles and a call with your pick before anything starts.
Placement runs through AYTalent. Engineers are employed and trained on the bench before they start with you.