A Win Has a Hundred Reasons - A Loss Usually Has One
A win can happen for a hundred reasons. A loss usually happened for one.
That’s an overstatement, but not by nearly as much as it should be. Wins carry a pile of contributing factors that all show up at once: good timing, a rep relationship that clicked, a competitor who went dark mid-evaluation, a champion who had effectively decided before the first demo. Great product and real execution matter too, and isolating which of those factors actually closed the deal is genuinely hard, because several of them are usually true at the same time.
Ask your own team why a specific deal was won and you’ll get an honest answer built on incomplete information. A rep remembers the moments they were in the room for: the demo that landed, the objection they handled well. What they can’t see is the champion making the internal case behind closed doors, or a competitor quietly stumbling on a requirement your team never knew was in play. A win gets credited to whatever was visible from your side of the table, which is rarely the same thing that was actually decisive.
Losses read differently. When a deal goes sideways, buyers can usually point to something specific: a demo that didn’t land for an identifiable reason, a pricing structure that surfaced a concern nobody on your team ever heard, a competitor who resolved an objection your messaging had no answer to. The signal is cleaner, and it’s exactly the kind of signal a program can act on.
This asymmetry is why every engagement I run is loss-heavy: twenty loss interviews against ten win interviews, out of a target range of twenty to thirty total. The losses are where I go hunting for the pattern, the recurring friction, the competitive narrative that won’t stop showing up. The wins do a different job. They test what the losses found.
Say a specific integration gap turns up in a dozen loss interviews, described independently by buyers who never spoke to each other. Run that same gap against the ten win interviews. If it never surfaces there, not even in passing, that’s a real, specific dealbreaker for a segment of buyers, and it needs a product or messaging response. If it does surface in a couple of the win interviews as something a buyer noticed and worked around anyway, the finding changes shape. It’s a real weakness, but not an urgent one, and the roadmap conversation that follows looks completely different.
When several things go right at once, you can’t isolate which one closed the deal. When one thing goes wrong the same way in interview after interview, you know exactly where to start.
That’s the entire logic behind weighting a program toward losses instead of splitting interviews evenly. A win-heavy program spends most of its budget chasing a signal that’s inherently harder to isolate, then treats whatever surfaces in the smaller loss set as an afterthought instead of the main event.
The mechanics of outreach point the same direction, which makes the win-heavy instinct even easier to fall into. Loss interviews convert at close to five percent, so reaching twenty completed conversations usually takes outreach to something like four hundred former buyers. Win participation rate runs close to double that, which makes win interviews faster to fill and tempting to lean on for exactly that reason. Easier to collect isn’t the same as more useful. A program that over-indexes on the win side because it’s the path of least resistance ends up with a stack of interviews that mostly agree with each other and confirm very little.
None of this makes win interviews optional. A pattern that shows up only in loss conversations and never once in a win conversation is a much stronger claim than a pattern from loss interviews alone, and a program that skips the win side has no way to make that comparison. The value of a win interview isn’t in what it reveals on its own. It’s in what it confirms, or fails to confirm, about what the losses already told you.
The ratio only pays off once the interviews get read as a set instead of a stack of one-off deal summaries. A recurring gap rarely announces itself in a single loss interview. It becomes a finding once it turns up again, unprompted, in the fifth conversation, and again in the twelfth, described in different words by buyers who never spoke to each other. Skip that comparison step and twenty loss interviews plus ten win interviews is just thirty transcripts, no sharper as a set than any one of them would have been alone.
If your next program splits interviews fifty-fifty because that feels balanced, you’re optimizing for the wrong kind of fairness. Weight it toward the losses. That’s where the answer usually is.