Less Free Time Sold More. Here Is Why: SaaS Free Trial Length

A shorter free trial sold more in one big test. Here is what really happens inside a long trial. And here is how to find the right SaaS free trial length for your own product. SaaS just means software people pay for each month.

Why would less free time make more people buy?

A big software company once ran a test on its free trial. It gave 337,724 new users a free trial. The length was picked by chance, like a coin toss.

Some people got 7 days. Some got 14. Most got 30.

The people with 7 days bought more.

Now think about your own trial.

You built the product on your own. You chose 14 days, or 30, because that is what everyone else does. Most software products pick 14 days. In one 2026 count of 200 products, 62 in 100 did (ChartMogul, February 2026).

People sign up. They click around for a day. Then you never hear from them again.

When the trial ends, almost nobody pays. In that same count, the usual product saw 8 in 100 free users become paying customers.

So you wonder. Did they need more time? Was the price too high?

The test above says neither of those may be the problem. So why would less free time make more people buy?

What you will know by the end

Most advice on trial length says "pick 14 days" or "it depends". Neither tells you what actually happens inside a trial.

By the end of this piece you will know:

• What 337,724 people did during their free trial, day by day.
• The usual reason given for long trials failing, and why the test showed it was wrong.
• When a longer trial really does sell more.
• How to find the right length for your own product, using numbers you already have.

Let us start with a library book.

What happened when 337,724 people got a free trial

You borrow a book from the library. You can keep it for 30 days.

You read a few pages on the first weekend. Then the book goes under the bed. On day 30, you have forgotten what it was about.

Keep that book in mind.

In 2023, three researchers published the results of a free trial test (Yoganarasimhan, Barzegary and Pani, Management Science, 2023). A free copy is on the lead author's website.

The test was run by a large software company. It was the leader in its field and sold a set of work programs. It kept its name private.

Before the test, every new user got a 30 day free trial.

During the test, new users in six places got 7, 14 or 30 days, picked by chance. The places were Australia and New Zealand, France, Germany, Japan, the UK and the US.

The company then watched those people for two years.

Here is how many bought, out of every 100 people:

• 7 day trial: 15.36 bought.
• 14 day trial: 14.96 bought.
• 30 day trial: 14.67 bought.

The 7 day trial won. It also kept customers paying longer, and it brought in more money over the two years. The 14 day trial was not clearly different from 30.

The gap is small. It is about 7 more buyers for every 1,000 people. But it points the opposite way to what most founders expect.

So what was going on?

Was it people using it for free and leaving?

There is a common answer to that. Many people believe a long trial lets people use the product for free, get their job done, and leave.

Think of someone who needs a design tool for one school project. Give them 30 free days and they finish the project. They never need to pay.

The researchers checked this idea. If it were true, the people who used the product the most would buy the least.

They found the opposite. People who used the product more during their trial were more likely to buy, not less.

So people taking it for free and leaving were not the reason. The usual explanation was wrong.

Before we get to what was really happening, it is fair to look at the other side.

When does a longer trial win?

A longer trial does not always lose. The same study found three cases where it did better.

First, experienced users. Some users already knew an older version of the software. They needed more time to compare it with what they had. For them, a longer trial worked better.

Second, some countries. Business users in Japan did better on the 30 day trial. Users in Germany did better on 14 days.

Third, simple products. The authors say a simple, easy product may do better with a longer trial. Their product was hard to learn.

There is also data from phone apps. A company called RevenueCat looks at many apps that people pay for each month. In 2026 it looked at trials of 17 to 32 days. On a typical app, 42.5 in 100 people paid. On trials under 4 days, 25.5 in 100 paid (RevenueCat, March 2026).

It also found that more than half of people on a 3 day trial cancelled almost at once.

That sounds like the opposite result. But it is not the same kind of test. RevenueCat compared different apps. Apps that choose long trials may simply be different apps. The 2023 study gave the same product different trials, picked by chance. That is a fairer test.

RevenueCat and ChartMogul both sell tools to companies that charge each month. Keep that in mind with both.

So longer can win. But in the fair test, it lost. Here is why.

What people did on day 10 of a 30 day trial

Go back to the library book.

The researchers looked at what people actually did during their trial. The average person used the product on only about 3 days.

It did not matter much how long the trial was. Going from 7 days to 14 days added well under one extra day of use.

Then they measured something else. How many days passed between the last time a person used the product and the end of the trial?

On the 7 day trial, that gap was 4.6 days on average.

On the 30 day trial, it was more than 21 days.

That is three weeks with the product sitting under the bed.

And the longer that gap, the less likely the person was to buy. The authors say a person who has not used it for a long time is likely to forget about it. Or they decide it is not useful.

One more detail. Take two people who used the product on the same number of days. The one on the shorter trial was more likely to buy.

So here is the answer. A longer trial did not give people more days of using the product. It gave them more days of forgetting it.

That leads to five things you can do this week.

How to find the right SaaS free trial length for your own product

1. Find the last day of use. Look at your last 20 trial users. For each one, write down the last day they used your product. Most tools show a "last seen" or "last login" date.
2. Count the empty days. For each person, count the days between that last day and the end of their trial. If most went quiet in week one of a 30 day trial, most of your trial is empty days.
3. Fit the trial to what you see. If most people stop by day 5, a 30 day trial mostly gives them time to forget. Try a shorter one. If you get enough sign ups, give half of new users the shorter trial for a month and compare.
4. Send one useful thing on the quiet day. Find the day most people go quiet. On that day, send one small task they can finish in five minutes. Not "your trial ends soon". This is our own idea from the study, not something the study tested.
5. Send a plain reminder before the trial ends. Say what happens next and what it costs. Do you sell to the public in the UK? Then this is the law from January 2027 (Latham and Watkins, September 2026).

One exception. Some users are experts moving from another tool. Some products are very simple. A longer trial may suit them. Your own last-day numbers will tell you.

What this study cannot tell you

• This was one company with one kind of software. It was the leader in its field. People may trust a big name and buy on that alone. A small, unknown product may see different results.
• The difference was small. About 7 more buyers per 1,000 people.
• The study shows that empty days went together with fewer sales. It does not prove the empty days caused it.
• The app numbers from RevenueCat compare different apps, not one app with different trials.
• ChartMogul and RevenueCat both sell tools to companies that charge each month. Their numbers come from their own customers.
• Step 4 is our own reading. Nobody has tested it in this study.

Frequently asked questions about free trial length

Is 7 days always the best free trial length?

No. It won in one big test of one product. Experienced users and simple products may do better with longer. Check your own last-day numbers first.

Should I ask for a card before the trial starts?

ChartMogul found trials that ask for a card see far more people pay, about 30 in 100. But fewer people may start the trial at all. That is a separate choice from length.

What if I only get a few sign ups a month?

You cannot run a fair test with a few people. Just look at the last day of use for each one. Even ten people will show you a pattern.

Does this apply to apps on phones?

Apps often take a card at the start and charge on the last day. That changes how people behave. The app data shows longer trials doing well there.

What counts as "using" the product?

Pick one action that shows real use. Opening the app is not enough. Saving or sending something is better.

One question for you

When people try your product for free, on which day do most of them stop using it?

Do you know that day?