Book a demo is winning on B2B software homepages

Product-led growth was the go-to-market story of the late 2010s: let people sign up, try the product and buy later. B2B software homepages went the other way. In 2019, 46.2% of readable B2B software homepages led with a demo or sales button such as "Book a demo" or "Contact sales". By 2026, 61.2% did. Over the same years, homepages leading with a self-serve button such as "Start free" or "Sign up" fell from 19.4% to 15.5%.

Among homepages that offer either path, demo and sales buttons went from 70.4% to 79.8% of the main calls to action.

What we measured

We followed a random sample of 10,000 B2B software companies from Ken internal data, headquartered in the US, the UK, Canada, Australia, Ireland or New Zealand, founded by 2018 and with 11 to 5,000 people today. For each year from 2019 to 2026 we read one homepage per company from Ken's web archive and found its main conversion button: the first demo or self-serve button in the top section of the page, or the one in the header if the top section has none. Between 3,207 and 4,175 homepages a year were readable.

Demo gained ground every single year

The demo share rose in every year from 2019 to 2026: 46.2%, 48.8%, 50.0%, 52.7%, 54.4%, 56.0%, 57.5% and 61.2%. It was the most common main call to action in every year.

This is not just new companies replacing old ones. Among the 645 companies whose homepage we could read in all eight years, the demo share rose from 49.0% to 60.0%, and the self-serve share fell from 24.5% to 19.5%. The same companies changed their buttons.

Two other things moved. Homepages with no clear conversion button, just "Learn more" or a menu, became rarer: 34.4% in 2019, 23.3% in 2026. Pages offering both paths side by side became a little more common, from 11.8% to 14.1%.

Every size moved the same way

The shift shows up at every company size, measured by headcount in each year. Among companies with 11 to 50 people, the demo share went from 52.5% to 68.2%. At 51 to 200 people it went from 59.2% to 73.8%. Even companies with 10 people or fewer moved, from 42.8% to 54.9%. The smallest companies lean least toward sales-led buttons, but the direction is the same everywhere.

Funding tells the same story. Companies with a dated raise on record went from 45.3% demo in 2019 to 63.5% in 2026, and companies without one from 46.4% to 60.3%. Funded companies also kept more self-serve buttons: 19.6% against 13.9% in 2026.

How sure are we about the size of the shift?

We checked our page reader against blind labels made without seeing its output, themselves not ground truth, and it is far from perfect. It agreed with them on 77% of 2023 to 2026 pages and 65% of 2019 to 2020 pages, below the 85% we set in advance. Its errors also drift: on older pages it misses some "Start free" buttons, and on recent pages it slightly over-reads "Book a demo".

Correcting for the errors we measured, the demo share rose from 46.0% to 56.8% rather than from 46.2% to 61.2%, and the self-serve share fell from 29.6% to 16.8% rather than from 19.4% to 15.5%. These corrections rest on small check samples, so treat them as rough. Either way, the direction holds: toward demo, away from free.

Did switching pay off? We can't tell

We also asked whether companies that changed their main button grew faster afterwards, and whether their sales team changed shape. Switches turned out to be rare. From 2021 to 2025, 104 companies moved cleanly from a demo button to a self-serve one, and 146 moved the other way.

After the strict matching we set out in advance, comparing each switcher with similar companies that kept their button, fewer than 50 switchers were left in each direction. Under our rules that is too few to report.

With looser matching, 70 demo-to-free and 61 free-to-demo switchers could be compared. Two years on, the median headcount gap was -2.06 points for demo-to-free switchers (95% interval -19.69 to 4.07) and -4.35 points for free-to-demo switchers (-16.67 to 9.52). There was no clear change in the number of SDRs either.

But this comparison fails a basic check: before the switch, switchers and their matches were not growing at the same pace. A slightly different way of choosing the comparison companies flips the free-to-demo result to +3.85 points. The data cannot say whether a button switch is linked with faster or slower growth.

What this data cannot show

• Survivors only. The sample is today's company records, so companies that closed before 2026 are missing, and early years lean toward survivors.
• Unreadable pages. About one homepage in seven is built by scripts we could not read, and those pages are left out.
• One page a year. The archive does not capture every site every year, and the 2026 reading comes from September, three months later in the year than the others.
• A button is not a business model. Industry labels also do not fully separate B2B from consumer software.
• No years before 2019. We expected self-serve to peak between 2019 and 2022. Within our window its highest year is 2019, so whether it peaked earlier is still open.

What this means for B2B teams

• The market has voted for a conversation. For most B2B software companies, the main ask on the homepage is a demo or a sales call, and that has become more true every year since 2019.
• Self-serve has not disappeared. About one homepage in six still leads with it, and pages that offer both paths are creeping up. Many teams now give buyers a choice rather than pick one model.
• Don't expect the button to do the work. Switches are rare, and we found no sign that companies grew faster after switching in either direction, though the data cannot rule it out.