Summary / TL;DR: How much Win / Loss analysis to do? About 20 to 40% of a well randomized sampling of all your deals on a quarterly basis.
When our clients develop a Win / Loss program, after a while, they want to get a sense of how many Win / Loss interviews they should be doing. This is a good question.
That one-off Win / Loss interview can be useful to satisfy that immediate craving for knowledge. “Why did we win?” or more often, “Why did we lose?” But, aside from that, the findings are typically not conclusive and, as a result, not actionable. Making wholesale changes on the back of the words of one interviewee is not something we want to do. This is because we don’t know whether we’ve just stumbled across a unicorn or if we have just discovered a thread that – if we pull on it – we will find woven through multiple deals.
That’s why when we talk about building a Win / Loss program, we talk about it like building muscle tone. We have to be consistent and rigorous in our application, and we recommend a well randomized sample of 20 to 40% of deals on a quarterly basis as the right frequency to build the desired muscle tone.
Of course, if significant changes are made in your sales motion, such as offer structure, product or brand marketing, you might want to temporarily increase range. Our aim is to keep the consistency and cadence out because we not only want to find those threads, pull them and see where they go, but – when we do identify an unfavorable pattern – we want to see the implementation of remediation plans that removes the unfavorable pattern.