We publish research on this site because it is the only evidence we can currently offer about the market. It would be a mistake to let it do more work than it can.

What an imported figure hides

A benchmark compresses a distribution into a point. When a study reports that buyers contact sellers at 61 percent of the journey, the useful information is not the 61. It is that the number exists on a distribution with firms at 30 and firms at 85, and that the distinguishing features between them are not in the headline.

Applied to a decision, the point estimate quietly becomes a target. A firm that adopts an industry average as a goal has committed to being average at that thing, which is rarely the intention and is almost never stated out loud.

Composition is the second problem. Most widely quoted business buying research is weighted towards technology purchases at larger companies, because that is who responds to these surveys. A founder-led services firm selling to owner-operators is not in that sample in any meaningful proportion.

What a baseline does instead

A baseline is measured on your business, at a stated date, using a method that can be run again. It produces a number that is comparable to itself over time, which is the only comparison that supports a decision about what to change.

It also surfaces the thing benchmarks cannot: which of the four surfaces is actually failing in your case. The published research says that offers are often unclear and that journeys are usually absent early. It cannot tell you that your offer is legible and your sales process is sound and the entire problem is that nobody encounters you during research. Only a measurement of your system does that.

How we use published research

Three ways, and we try to keep them separate. To establish that a phenomenon exists at all. To decide which direction to look first. To argue with our own assumptions when a study says something we did not expect.

We do not use it to size an opportunity in a specific business, to set a target, or to support a claim about what an engagement would produce. Those require a measurement of the business in question, and we would rather say that than fill the gap with an average.