Every fall, we publish "The State of B2B SaaS 2026," a report with a confident title, a gradient cover, and a methodology section most readers never scroll to. We'd like to walk through how it's actually made, because the gap between the title and the methodology is where most of these reports quietly live.

The sample size

We surveyed 203 people. The title does not say "according to 203 people who happened to open our email." It says "The State of B2B SaaS," definite article, no hedge, as though 203 responses constitute a census of an entire industry rather than a reasonably sized focus group that skews heavily toward people who already subscribe to our newsletter and therefore already agree with several of our product's core assumptions.

Who actually responded

The survey went to our existing email list, which means every respondent was already a customer, a prospect, or someone who downloaded a gated whitepaper at some point and got added to a drip campaign. This is not a representative sample of "B2B SaaS." It's a representative sample of people already inside our funnel, asked questions partly designed to produce answers that make our own product category look urgent.

The leading questions

"How significant is AI adoption to your team's strategy in the next 12 months?" has five answer options, and all five options describe AI adoption as happening to some degree — there is no option for "we have not thought about this and are not going to." The survey can't produce the finding "most companies aren't prioritizing AI" because the question was never built to allow that answer to exist.

The stat that becomes a headline

203 respondents, a plurality of 61 picking one answer, becomes "30% of B2B SaaS companies say AI is their top priority," a sentence that sounds like an industry-wide finding and is actually a plurality of a email list skewed toward people who already use AI-adjacent language in their job titles. The percentage is real arithmetic. The population it's being generalized to was never surveyed.

Why this report exists

Not primarily to inform the industry — to generate a piece of content that trade publications, LinkedIn influencers, and other companies' own blog posts will cite as "research shows," each citation linking back to us, each one lending the number a little more borrowed authority the further it travels from the methodology section nobody read. A stat with our logo on it, repeated by someone else, becomes evidence of a trend rather than a marketing artifact, purely by being quoted enough times that its origin stops mattering.

The "in partnership with" version

Sometimes two companies co-publish the same kind of report, which reads as more credible — two logos implies two independent perspectives converging on the same finding. In practice, it usually means two marketing teams split the cost of the same survey tool and agreed in advance on which three findings would make the press release, because a joint report with two logos gets covered by more outlets than a report with one, for reasons that have nothing to do with rigor.

What we're not saying

We're not saying every industry report is fake — some are run by real research teams with real sampling discipline, large enough and random enough to say something true about an actual market. We're saying "State of the Industry," unqualified, is often doing the same thing "leading operations team" does in a case study: borrowing the authority of a category the data was never actually sized to speak for.

Our recommendation

Keep running the survey — it's genuinely useful content, and your own customers' opinions are worth publishing. Just title it honestly: "What 203 of Our Subscribers Think About AI Adoption" doesn't travel as far as "The State of B2B SaaS 2026," but it's the sentence that's actually true, and the sentence that's actually true survives a skeptical reader finding the methodology page.

If your last industry report surveyed 200 of your own subscribers and titled itself "The State of" an entire category, email us at contact@b2bsaas.guru. We'll help you write next year's. We'll probably keep the confident title. It gets cited more.