# Half of B2B software homepages now say AI. Did it pay off?

49.0% of B2B software homepages mentioned AI in 2026, up from 11.7% in 2019. Companies that made AI their main claim grew headcount no faster than matched peers afterwards.

By Cristian Frunze, Founder of Ken · Oct 16, 2026 · [Ken Research](/research) No. 9 · Websites and pricing, AI and work, Founders and startups

## Key findings

- **49.0%** of B2B software homepages mentioned AI in September 2026, up from 11.7% in 2019 and 15.8% in 2022, the year ChatGPT launched.
- **25.8%** made AI the main claim, in the page title, description or main headline, up from 4.6% in 2019 and 5.3% in 2022.
- **-8.2%** headcount gap after eight quarters for companies that made AI their main claim, against matched peers (95% interval -16.9% to 1.3%).
- **9.3% vs 8.4%** of adopters and matched companies raised money in the 24 months from the start of the change year. The difference is within the noise.

Half of B2B software companies now put AI on their homepage. In September 2026, **49.0% of B2B software homepages mentioned AI**, up from 11.7% in 2019. For 25.8%, AI was the main claim, up from 4.6%.

It did not come with faster growth. Companies that made AI their main homepage claim grew headcount no faster over the next two years than matched companies that didn't, and if anything trailed them: after eight quarters the gap was -8.2% (95% interval -16.9% to 1.3%). This is a link, not proof of cause.

**Figure 1. AI mentions were flat until ChatGPT, then took off**

Share of B2B software homepages that mention AI anywhere, and that make AI the main claim (title, description or main headline), one crawl per year. Bands are 95% intervals. The 2026 point is a September crawl; the others are June or July.

| Crawl year | Mentions AI | Mentions AI (95% interval) | AI as the main claim | AI as the main claim (95% interval) |
| --- | --- | --- | --- | --- |
| 2019 | 11.7% | 10.8% to 12.6% | 4.6% | 4.0% to 5.3% |
| 2020 | 14.0% | 13.0% to 15.0% | 5.3% | 4.7% to 6.0% |
| 2021 | 15.4% | 14.4% to 16.4% | 5.7% | 5.1% to 6.4% |
| 2022 | 15.8% | 14.8% to 16.9% | 5.3% | 4.7% to 6.0% |
| 2023 | 19.3% | 18.2% to 20.5% | 6.9% | 6.2% to 7.7% |
| 2024 | 26.3% | 25.1% to 27.6% | 10.7% | 9.8% to 11.6% |
| 2025 | 33.0% | 31.7% to 34.4% | 15.2% | 14.2% to 16.3% |
| 2026 | 49.0% | 47.4% to 50.6% | 25.8% | 24.4% to 27.2% |

*Source: Ken internal data (Ken's archive of company websites and the company and career records Ken uses for sales and outreach), analysed by Cristian Frunze.*

## What we measured

We drew a random sample of 10,000 B2B software companies from Ken internal data, with headquarters in the US, the UK, Canada, Australia, Ireland or New Zealand and 11 to 5,000 employees. Then we read each company's homepage once a year in Ken's web archive, from 2019 to 2026. Each year we could read between 3,704 and 4,847 of them.

A homepage **mentions AI** if it uses AI, artificial intelligence, machine learning, LLMs, generative AI, GPT or AI agents anywhere on the page. **AI is the main claim** when it appears in the page title, the description or the main headline, or three or more times near the top. A language model double-checked the unclear cases. On a random sample of 300 homepages it agreed with the word rules 98.0% of the time on mentions and 96.3% on main claims.

## Flat for four years, then a sharp bend

For four years the share barely moved: 11.7% in 2019, 15.8% in 2022. Then ChatGPT came out at the end of 2022 and the line bent. The share reached 19.3% in 2023, 26.3% in 2024, 33.0% in 2025 and 49.0% in September 2026. AI as the main claim followed the same path, from 5.3% in 2022 to 15.2% in 2025 and 25.8% in 2026.

This is not just new companies joining the sample. Among the 878 companies whose homepage we could read in all eight years, the share that mentioned AI went from 13.4% to 52.2%.

We predicted that by 2026 more than half of homepages would mention AI and about a quarter would make it the main claim. The main claim landed on a quarter. Mentions ended just under half, and part of the last jump is timing: the 2026 reading comes from a September crawl, about 15 months after the June 2025 one.

## Did it pay off in growth?

Growing companies tend to change their websites, so comparing AI adopters with everyone else would flatter AI. Instead we compared each company that first put AI on its homepage with up to three similar companies that did not: same industry group, same founding period, same size, and the same headcount growth over the two years before.

Before the change the two groups grew alike. For the main-claim group, the largest gap in the eight quarters before was 0.9%, and a joint test found no difference (p = 0.597).

**Figure 2. Adopters grew like their matches before the change, and no faster after**

Headcount of companies that made AI their main homepage claim, minus their matched companies, in log points (small values read roughly as percent). Quarter 0 is the last quarter of the year before the change. The band is the 95% interval. 260 adopters up to quarter 5, 140 after.

| Quarters from the last quarter before the change year | AI as the main claim, against matched companies | 95% interval |
| --- | --- | --- |
| -8 | +0.0042 | -0.002 to +0.0105 |
| -7 | -0.0003 | -0.0135 to +0.0136 |
| -6 | +0.002 | -0.0129 to +0.0178 |
| -5 | -0.0002 | -0.0121 to +0.0117 |
| -4 | +0.0036 | -0.0007 to +0.008 |
| -3 | +0.0047 | -0.008 to +0.0169 |
| -2 | +0.009 | -0.0059 to +0.0236 |
| -1 | +0.0078 | -0.0031 to +0.0184 |
| 0 | 0 | 0 to 0 |
| +1 | -0.0013 | -0.0154 to +0.0133 |
| +2 | +0.0049 | -0.0174 to +0.0281 |
| +3 | -0.0069 | -0.0407 to +0.0221 |
| +4 | -0.0128 | -0.0547 to +0.0232 |
| +5 | -0.0147 | -0.0633 to +0.0269 |
| +6 | -0.0341 | -0.1178 to +0.0413 |
| +7 | -0.0626 | -0.157 to +0.0227 |
| +8 | -0.0855 | -0.1872 to +0.0059 |

*Source: Ken internal data (Ken's archive of company websites and the company and career records Ken uses for sales and outreach), analysed by Cristian Frunze.*

After the change, the 140 companies that made AI their main claim and could be followed for eight quarters had a headcount gap of -8.2% against their matches. The interval runs from -16.9% to 1.3%, so this is not a clear decline, but it leaves little room for a growth payoff. Medians are milder, a gap of -2.6 points, because companies that shut down count as zero headcount and pull the average down. The 340 companies that only mentioned AI had a gap of -3.4% (95% interval -8.7% to 1.7%).

**Figure 3. Every headcount gap includes zero, and most lean negative**

Headcount growth of adopters minus matched companies, after 4 and 8 quarters, for both definitions. Dots are mean gaps converted to percent; lines are 95% intervals.

|   | AI as the main claim | 95% interval | Mentions AI | 95% interval |
| --- | --- | --- | --- | --- |
| After 4 quarters | -1.3% | -5.1% to +2.7% | -2.1% | -4.9% to +0.4% |
| After 8 quarters | -8.2% | -16.9% to +1.3% | -3.4% | -8.7% to +1.7% |

*Source: Ken internal data (Ken's archive of company websites and the company and career records Ken uses for sales and outreach), analysed by Cristian Frunze.*

Funding looked the same. 9.3% of main-claim adopters raised money in the 24 months from the start of the change year, against 8.4% of matched companies, a gap of 0.8 points (95% interval -5.4 to 6.4).

One outcome leaned the other way. On the broad definition, 10.9% of adopters had no readable homepage in the 2025 or 2026 crawls, against 16.4% of matched companies. That interval also includes zero (-11.1 to 0.6 points).

We also expected adopters to have been growing faster before they changed their homepage. They weren't: before matching, their prior growth equaled that of the median non-adopter in the same year. Hot companies were not the ones reaching for AI.

## What this data cannot show

- **Correlation, not cause.** The matched comparison rules out that adopters were already growing faster, but not that something else changed at the same time, such as a pivot or a struggling business reaching for a new story.
- **Mostly the recent wave.** Most matched adopters changed in 2024 and 2025. The result describes the generative AI rush, and a payoff that takes longer than two years would not show yet.
- **Headcount, not revenue.** Headcount is rebuilt from work histories of people in six English-speaking countries, and the latest quarters undercount for everyone. Both groups are measured the same way.
- **Words, not products.** A homepage can say AI without the product using it. The two definitions bracket this.
- **Survivors and coverage.** The sample comes from today's company records, so companies that disappeared before 2026 are missing. The archive reads about half of these homepages in any one crawl, and industry labels do not fully separate B2B from consumer software.

## What this means for B2B teams

- **AI on the homepage is the default, not a differentiator.** About half of B2B software homepages say it. Saying it too won't set you apart.
- **Don't expect the claim to move growth.** Companies that made AI their headline grew no faster over two years than matched peers, and raised money about as often.
- **Back the claim with the product.** If AI does real work in your product, show what it does for the buyer. A vague mention puts you in a crowd of 49.0%.
- **Read prospects' homepages with this in mind.** An AI headline no longer tells you much about a company's product, budget or momentum.

## Data and method

- **Data:** Ken internal data (the company records and work histories Ken uses for sales and outreach, reused by Cristian Frunze for this study), with one homepage a year from Ken's web archive. Funding from SEC Form D filings and the last round on company records.
- **Sample:** 10,000 B2B software companies with 11 to 5,000 employees in six countries, 3,704 to 4,847 readable homepages a year. 260 matched adopters on the main-claim definition, 500 on the broad one.
- **Date range:** Homepages from June 2019 to September 2026. Headcount from 2015 Q1 to 2026 Q1.
- **Method:** Pre-registered. Word rules with a language-model check label each homepage. Adopters matched to up to three companies on industry, founding period, size and prior growth, with an event study and bootstrap 95% intervals. Headline numbers recomputed by a second, independent analysis.
- **Excluded:** Parked, for-sale and placeholder pages; Records whose website domain is shared with another company, or is a platform domain; IT services and consulting firms, which mostly sell services rather than software.
- **Privacy:** Aggregates only. Any group smaller than 50 people is suppressed, and no individual is named.

[Full methodology as Markdown](/research/ai-homepages/methodology.md) · [Download the full methodology (PDF)](/research/ai-homepages.pdf)

## Related studies.

- No. 1: [Did AI kill the SDR role?](/research/did-ai-kill-the-sdr-role.md) - Sep 30, 2026 · AI and work, Sales teams. 6.7% of software sales teams were SDRs in March 2026. Not in headcount. SDRs were 6.7% of software sales teams in March 2026, against 6.8% a month after ChatGPT launched. Full methodology: [Markdown](/research/did-ai-kill-the-sdr-role/methodology.md) · [PDF](/research/did-ai-kill-the-sdr-role.pdf).
- No. 10: [Where does Series A money go? As much to selling as to building](/research/series-a-hiring.md) - Oct 18, 2026 · Founders and startups, Hiring and org design, Sales teams. 26.9% of Series A net adds go to selling, level with building. Into more people, split evenly. In the year after a Series A, software companies put 27.0% of net adds into building and 26.9% into selling, and added 13.4 people each, against 7.5 the year before. Full methodology: [Markdown](/research/series-a-hiring/methodology.md) · [PDF](/research/series-a-hiring.pdf).
- No. 7: [The 18-month VP of Sales is a myth: the median is 32 months](/research/vp-of-sales-tenure.md) - Oct 12, 2026 · Sales teams, Hiring and org design, Founders and startups. 32 months median tenure of a VP of Sales, CRO or Head of Sales. The median VP of Sales, CRO or Head of Sales stays 32 months, and 72.4% are still in the seat at 18 months. When the only sales leader leaves, the company falls about 10% behind its peers within a year. Full methodology: [Markdown](/research/vp-of-sales-tenure/methodology.md) · [PDF](/research/vp-of-sales-tenure.pdf).

## Put the data to work.

No card · Or we run it for you

[Get early access](https://app.ken.so/auth/sign-up)