AI reshaped how products are built, priced, and sold. Self-serve funnels now include AI-driven onboarding and in-product copilots. Supporting free users has gotten much more expensive as AI token costs have remained high. AI is even changing who signs up for and uses products in the first place, from people to a combination of people and agents.

What hasn’t changed: improving free-to-paid conversion remains one of the highest-leverage ways to grow revenue, without adding headcount or increasing marketing spend.

To understand what good looks like in 2026, I partnered with my friends at ChartMogul and ProductLed to analyze conversion data from 200 B2B software products.

Key stats from the report:

This is the first report in a two-part series; subscribe to receive the next report when it’s released in a few weeks.

The study was conducted in January 2026 and included 200 software products. A typical respondent is between $1 to $10 million ARR with an average revenue per customer of $50-$249 per month and a year-on-year growth rate between 25-50%.

The survey defined free-to-paid conversion as the percentage of leads or free signups that convert to become a paying customer within six months.

Respondents by annual recurring revenue (ARR):

Respondents by average revenue per customer (in USD):

Respondents by ARR growth rate:

Respondents by product type:

Free trials are the overwhelmingly most popular entry point for the products in the research. 57% have a free trial, more than twice the rate of freemium (26%).

Just 7% have a reverse trial, i.e. temporary access to premium features while the user is on a free plan. Another 7% have an interactive demo experience where users get a hands-on product experience with dummy data. The rest (4%) have a paid trial, i.e. time-limited access with a special introductory price.

SaaS products seem particularly fond of free trials. 61% of SaaS products have a free trial as their primary landing point for new customers. This is higher than products that are a hybrid of SaaS and AI (51%) or AI-native (43%).

Among free trial products, 20% require users to provide a credit card upfront and the remaining 80% do not.

The most common trial length is 14 days (62% of products) followed by 7 days (14%) or 30 days (14%). These aren’t purely self-serve experiences; 80% of free trial products have human touchpoints when an enterprise user enters the trial.

Among freemium products, 38% let users try the product before creating an account like you might’ve seen with vibecoding tools (ex: Lovable, Replit) or LLMs (ex: ChatGPT, Perplexity). 70% of freemium products have human touchpoints when an enterprise user self-serves.

The free-to-paid conversion rates across these products vary substantially. I’m not talking about a percentage point or two. There’s a 10x conversion difference between the top 20% of self-serve products and the bottom 20%!

The median free-to-paid conversion rate across all products was 8%. But very few products actually have an 8% conversion rate, as you can see in the chart.

Here’s what the data shows for free trial products:

By comparison, here’s what the data shows for freemium products:

On average, 3%-5% is a GOOD conversion rate for a freemium self-serve product, and 8%-12% is GREAT. For a freemium product with an ungated experience, i.e. where a user can experience the product without creating an account, a GOOD conversion rate is 7%-9%.

Products that lead with a free trial have slightly higher conversion rates on average – GOOD is 4-6% and GREAT is 10%-15%.

Reverse trials performed somewhere in between free trials and freemium. GOOD conversion is 4%-6% and GREAT is 8%-12%. The sample size here was relatively small; this difference is not statistically significant.

Free trial products that require a credit card see substantially higher conversion compared to ones that don’t. GOOD is 25%-35% and GREAT is 50%-60% when a credit card is required. (It’s worth mentioning that introducing a credit card requirement could blunt signups, hurting the overall number of paying customers, unless the product warrants the friction of adding a credit card. More on that in the next section.)

AI-native and AI plus SaaS hybrid products see slightly higher conversion compared to traditional SaaS. GOOD is 6-8% and GREAT is 15-20%. This is notable because AI-native products are more likely to be freemium compared to SaaS products. Median conversion is higher despite the mix shift.

Similarly, B2B-focused products see somewhat higher conversion than B2C or hybrid B2C/B2B products. (22% of the products in the survey were either mostly B2C or a hybrid of B2C and B2B.)

Free-to-paid conversion is only half of the equation. You can only convert users who sign up in the first place. A miserly free experience could turn off your audience, or cause them to wait until they’re more serious about adopting your product – hurting overall signups and customer growth.

Let’s layer in signup rates, or the percentage of website visitors who create an account in the first place.

For every 1,000 website visitors:

What the data tells me: freemium or free trial is an outdated question. Neither provides the best chance to drive product adoption OR to maximize conversion.

If your goal is product adoption, you’re likely better off opening up the product and letting people interact with it before creating an account.

I’ve seen as much as a 3x increase in the number of website visitors who start using an ungated product, although slightly fewer of them create an account compared to a classic freemium experience. What good is capturing an email address if the person has written-off your product? (Plus, you could probably guess the person’s email address with a website de-anonymization tool if it was that important.)

If your goal is conversion, you’re likely better off going in the opposite direction and requiring a credit card in order to begin a trial. I’ve personally seen this approach from the likes of Shopify, Canva, Grammarly, AWS, Hootsuite, Fyxer (profiled earlier), Google Workspace, and many others. For more serious products or ones that touch sensitive information, your users might even expect a credit card requirement.

The best companies earn buyer trust by providing transparency about how and when users can cancel. Canva, for example, tells prospective trial-ers that they’ll get a reminder 7 days before their Canva Pro trial ends, de-risking the decision to sign up. Companies that don’t do this risk fewer trial signups, worse month 1-3 retention, and higher refund rates as people forget to cancel.

If you only take one thing from this report: stop copying another company’s free model without understanding what it optimizes for. Before changing your onboarding, pricing, or trial structure, answer these three questions:

The worst middle ground is a timid free experience that neither drives adoption nor creates urgency to buy.

In the next report, we’ll share how teams move these numbers: what channels convert best, who owns self-serve conversion internally, when sales gets involved, and which tools can help. The follow-up report will be released in the coming weeks – subscribe to get it delivered straight to your inbox.

Related resources: