Free trial conversion rate is the share of SaaS trial signups that become paying customers. It is simple to calculate and almost always quoted without the context that makes it meaningful.
Benchmarks, with the caveat that matters
The single biggest driver of the number is whether you ask for a credit card up front.
- No card required — conversion typically runs 8–15%. Signup volume is high and intent is mixed; a meaningful share of signups are researchers, students and competitors.
- Card required — conversion typically runs 40–60%. Volume drops sharply, often by two-thirds, but everyone who signs up has already decided to buy unless something goes wrong.
Comparing these two numbers directly is meaningless. A product moving from no-card to card-required will see conversion jump from 10% to 50% and may well end up with fewer customers. The metric that matters is paying customers per thousand visitors, not conversion rate.
Which model to choose
Card-required suits products with a short time to value, an established brand and a clear buyer. It filters hard, which is efficient when your sales or support capacity is the constraint.
No-card suits products that need exploration time, sell into teams where the evaluator is not the buyer, or depend on viral spread. It also produces a much larger pool of signups to market to, which has value even at low conversion.
A middle option that works well: no card for the trial, but card required to invite teammates or export data. The commitment arrives at the moment the user has already decided the product is useful.
Trial length
Fourteen days is the most common choice and is right for most products. The reasoning is behavioural rather than technical: nearly all trial activity happens in the first three days and the last two, whatever the length.
A 30-day trial mostly adds three weeks of nothing happening, and it pushes your revenue a month out. Seven days works for simple products with immediate value. Longer than 30 days is only justified where the product requires data to accumulate or an implementation to complete.
More effective than extending the length is extending on request. A one-click “need more time?” offer in the final-days email converts well and tells you exactly who to talk to.
Where trials actually fail
Conversion is usually lost in the first session, not at the paywall. The typical funnel looks like this:
- Signup to first login: 10–20% never return after creating the account.
- First login to activation: the largest drop. Half or more never reach the action that demonstrates value.
- Activation to habit: users who activate but do not return in week one rarely convert.
- Habit to purchase: comparatively small drop. Users who got value and came back mostly buy.
This shape is consistent enough to be useful: if your conversion is poor, the problem is almost certainly in step two, and pricing-page or paywall optimisation will not touch it.
Six fixes, in order of impact
1. Define and instrument the activation moment
Identify the specific action that correlates with retention — the first report run, the first integration connected, the first teammate invited. Measure the share of trials reaching it. Everything else on this list is downstream of knowing this number.
2. Remove setup from the critical path
Every field in the signup flow costs conversion. Company size, role, phone number and “how did you hear about us” can all be collected later, or inferred. Ask for an email and a password, and get the user into the product.
3. Start with data in the product
An empty state is the most common place trials die. Sample data, a template, or an import that pulls real data in one click all let the user see the product working before they have invested effort. This single change often moves activation by double digits.
4. Send behaviour-based emails, not scheduled ones
A day-three email that says “here’s what you can do next” is ignored. An email triggered by what the user did or failed to do — “you connected your data but haven’t built a report yet, here’s the two-minute version” — performs several times better. Our notes on welcome email sequences cover how to structure this.
5. Make the end of the trial a soft landing
A hard cut-off that locks the user out of their own work creates resentment. A read-only mode that preserves everything and makes upgrading a single click converts better and leaves the door open for the people who come back in two months.
6. Call the accounts that matter
In a product with meaningful contract values, a human reaching out to trials from target-profile companies lifts conversion far more than any automated sequence. Use firmographic data to pick the twenty worth a conversation rather than emailing all two hundred.
Measuring it without fooling yourself
Three adjustments keep the number honest. Exclude obviously invalid signups — disposable email domains, your own team, competitors you can identify. Measure by cohort rather than in a calendar month, so a trial that started on the 28th is not counted as a loss. And segment by acquisition channel: a blended 11% that is 25% from organic search and 4% from a paid campaign is telling you to change the campaign, not the product.
A worked example of where the money goes
Numbers make the funnel argument concrete. Suppose 2,000 people visit the pricing page in a month, 300 start a no-card trial, and 30 convert — a 10% trial conversion rate.
Breaking the 300 down: 45 never log in again after signup. Of the 255 who do, 140 never reach the activation milestone. Of the 115 who activate, 62 do not return within seven days. That leaves 53, of whom 30 buy — a 57% conversion from the habit stage.
The headline number says 10%. The funnel says the bottleneck is the 140 people who logged in, looked around and never got the product to do anything useful. Doubling conversion does not require touching pricing, the paywall or the trial length; it requires getting 70 more of those 140 to their first result.
This is why teams that optimise the end of the trial — discount offers, extension emails, urgency timers — usually see small gains. They are working on the stage where conversion is already 57%.
What to do with the ones who do not convert
Most companies treat an expired trial as a loss and stop. Three of those groups are worth separating.
Activated but did not buy. These are the most valuable non-customers you have. They got the product working and chose not to pay, which means price, timing or a missing feature. Ten phone calls here will tell you which, and the answer is usually specific enough to act on.
Never activated. A re-engagement sequence works poorly; a product fix works well. Treat this group as a research sample rather than a marketing list.
Wrong fit entirely. Students, competitors, people from companies you do not sell to. Filtering these out of your conversion denominator will usually improve the reported rate by a point or two and, more importantly, stop you optimising for them.
A quarterly win-back to the first group — particularly when you ship the thing they said was missing — reliably outperforms generic re-engagement campaigns.
Frequently asked questions
Should I require a credit card for a free trial?
Only if time to value is short and your brand is known enough that visitors will give you a card. Otherwise the volume loss usually outweighs the conversion gain. Measure paying customers per thousand visitors to decide.
Does extending a trial help?
Rarely as a default, often as an exception. Extending on request converts well because the request itself is a buying signal.
What counts as the start of a trial?
First login, not account creation. Counting signups who never logged in makes every downstream number look worse and hides the fact that the problem is email confirmation.
Is a reverse trial better?
A reverse trial — full access for a period, then automatic downgrade to a limited free tier — works well for products with a credible free tier. It combines trial urgency with the retained audience of freemium, and it avoids the hard cut-off that drives people away.
