The optimal free trial length

In a large-scale experiment, 7-day long free trials (vs 30 days) led to 5.6% higher conversions, 6.4% better retention, and 7.9% higher revenue.

October 11, 2022

Topics: Brand & Strategy | Promotions
For: B2C, B2B
Research date: August 2022
Universities: Foster School of Business, University of Washington

New to Ariyh? This is a 3min practical summary of a scientific study 🎓 
Subscribe for $0 to get a new science-based marketing insight every week 📈

📝 Intro

Free trials let people try a product for free for a limited time (e.g. “Try for free for 14 days”). They can then pay to stay, or leave.

This acquisition strategy is especially popular for SaaS businesses, which charge a subscription for their software (e.g. Netflix video streaming, Shopify ecommerce software, Hootsuite social media management). It’s often an alternative to a freemium business model.

But how long should free trials be?

The vast majority of trials offered by companies range between 7 days and 90 days, with most concentrated in the 7 to 30 days bracket.

Long trials give people more time to learn and use the product. They might make products stickier by making it harder to switch back.

Short trials create an urgency effect and might reduce the number of people that free-ride without ever buying.

A new study from the University of Washington gives us the answer.

📈 Recommendation

Shorten your free trial to 7 days. People will be more likely to convert and stay in the long term.

Personalized trial lengths (e.g. longer trials for experienced users, shorter trials for beginners) could boost your results further, but the increase is small and they are complex and risky to implement.

🎓 Findings

  • 7-day free trials (vs 14-day and 30-day) maximized acquisition, retention, and profits.

  • A large-scale field experiment tested 3 trial lengths with 337,724 users of a major SaaS company (anonymous but similar to Microsoft 365, Dropbox, or Adobe). The results found that:

    • 7-day trials increased

      • Subscriptions by 5.59%

      • Retention by 6.4% (2 years after)

      • Revenue by 7.91%

    • 14-day trials were not statistically different from 30-day trials

    • 30-day trials were the previous standard and were used as the baseline for comparison

  • Researchers also explored a method of giving personalized free trial lengths. Compared to a 7-day trial for all, the benefit of personalization was small (6.8% vs 5.6%). It was also complex and had a high rate of failure in many of the statistical models tested.

🧠 Why it works

  • The more intensively we use a product during a free trial the more likely we are to buy it.

  • Longer trials give us more time to learn so we should be more likely to convert.

  • However, when we are inactive during a trial - especially during the last days - we are less likely to buy.

  • And a lack of urgency of long trials means that we feel less pressure to use the product and are more likely to forget about it.

  • So while our total product use increases in long trials, we use it much less per day on average, especially in the last crucial days.

  • The result is that a shorter, but more intense, use of the product in a short trial leads to higher conversions.

💻 Brought to you by Storyblok

Want a faster, more effective way to run your website?

Try Storyblok - the next generation headless CMS:

  • Edit your website as you browse it

  • Launch new landing or product pages in minutes

  • Create content once, publish across all your channels

Try it out yourself for free - no credit card required.


✋Limitations

  • The experiment only tested 3 lengths of free trials (7, 14, and 30 days). It’s unclear what the effect of other lengths is, especially when they are outside this range (e.g. 3 days, 90 days).

  • The study is based on one field experiment in one product category. This makes it riskier to generalize the results to all contexts. Test it in your context.

  • User free riding was not an issue in this context, but could be for some products that are especially suited for one-off uses (e.g. survey software).

  • Although it was not a goal of this study, we do not have a comparison of the effect of having a free trial vs not having one at all.

🏢 Companies using this

  • Microsoft 365 and NordVPN offer 30-day free trials.

  • Spotify and YouTube Music offer trials that vary from 30 to 90 days depending on changing offers or the channels people start them from (e.g. customers of partner products get longer trials).

  • Marketing software HubSpot offers a 14-day trial.

  • Clio legal management software offers 7-day trials.

⚡ Steps to implement

  • Update your free trial to 7 days if you think it fits with your product offering.

  • Depending on your context, you could experiment with further reducing the length to see if it causes positive effects (e.g. 24 hours free trial).

🔍 Study type

Field experiment (with a large anonymous SaaS company between 1st December 2015 and 6th January 2016)

📖 Research

Design and evaluation of optimal free trials. Management Science (August 2022).

🏫 Researchers

Remember: This is a new scientific discovery. In the future it will probably be better understood and could even be proven wrong (that’s how science works). It may also not be generalizable to your situation. If it’s a risky change, always test it on a small scale before rolling it out widely.

🎓 Found this insight useful?