What the benchmarks actually say (and why averages mislead)

The benchmarks are useful exactly once you stop treating them as a target. The spread across B2B is enormous, and the spread is the lesson. In FirstPageSage's industry data, legal services convert at around 7.4% while B2B SaaS sits near 1.1%, the highest and lowest of the set. That is roughly a sevenfold gap between two legitimate B2B businesses, and it is not because legal firms build better landing pages. It is buyer intent and sales complexity. Someone searching for a solicitor often has an acute, urgent problem and a short consideration cycle; someone evaluating a SaaS platform is comparing five tools across a six-month committee process. The page barely moves that.

The gap that is offer and intent, not cosmetics

The most instructive number in the whole space is the distance between median and top performer within a single category. In B2B SaaS, the visitor-to-lead median runs roughly 1.5 to 2.5%, while top performers hit 8 to 15% on demo-request forms, a 3 to 5x difference inside the same industry. If the variance were cosmetic, that gap would close with a better button colour. It does not, because the gap is structural. Top performers earn higher-intent traffic, match a sharper offer to that intent, and present a call to action a ready buyer actually wants. They are converting a better-qualified visitor, not converting the same visitor harder.

This is why I distrust the instinct to benchmark against a cross-industry median. A widely cited figure around 2.9% is sometimes treated as the line every B2B site should clear, but it is noise the moment you remember SaaS sits at 1.1% and legal at 7.4%. A SaaS founder beating themselves up for missing 2.9% is measuring against a number their entire category structurally cannot reach on cold traffic.

How to find your real benchmark

Segment, do not compare to a headline. Your real benchmark is built from two cuts. First, your industry: find where businesses with your sales complexity and buyer intent actually land, and treat that band as your gravity, not the cross-industry average. Second, your own best source: if your highest-intent traffic, usually paid search or branded organic, is converting well below what that intent should produce, you have a capture problem you can fix, regardless of where your blended number sits.

The practical test I apply: take your single highest-intent channel and ask whether it converts like high-intent traffic. If your paid-search visitors, people typing your category into Google and clicking your ad, convert under 3%, something in the capture layer is leaking, because that audience is as close to the in-market 5% as you will ever buy. If they convert at 5% or better but your blended rate is 1.4%, the page is fine and your problem is upstream in your channel mix. Two founders with the same blended number can have completely opposite problems, and only segmentation tells them apart.