The Effect of Shorter Trial Periods for SaaS

I reduced my trial period for Planscope from 30 days to 14, and have already seen a dramatic uptick in sales. Unless you're doing something that takes a long time to really shine (like KISSmetrics), 14 days is more than enough to demonstrate how things work and see if there's a fit.

I wonder if part of the reason is that on shorter trial periods, people are still in the 'honeymoon phase'. When you make a good first impression, people will go 'oh this is the best thing ever'. Nothing ever is, of course, and it takes a while (weeks to months, depending on the product) to find out that some aspects aren't that great. But when you've made the first payment, you've committed yourself, and now you're prone to confirmation bias.

After 2 months, people are more likely to have hit the not-so-great parts and may think 'meh, I'm going to try other options first'. The shorter the trail period, the more people (at least of those who actively try at all) will still be in the honeymoon phase and more likely to sign up.

Just thinking out loud, I have no numbers to back this up, and I wonder how one could measure this. I guess you'd need to correlate user activity to the duration of the trial and the likelihood that somebody signs up - maybe there is an overarching trend to be found there somewhere. It would require a fairly homogeneous set of users though, or otherwise a large number of users and additional data (either user traits or usage pattern commonalities) to slice the users into more homogenous groups.

I'd love to see you perform some more tests (10, 20, 30) now that you have settled into your new website - for the good of the Internet!

Mileage, I'm sure, will vary depending on the audience.

Our company serves organisations of all sizes. I reckon a 14 day trial period would improve (at least the immediacy of) our conversions of small business customers, but it would simply cause difficulties and annoy our enterprise customers, often with a team of people involved at varying levels who are often spread over the world, resulting in a slow ramp-up to testing.

Targeting different trial lengths at different stereotypes of lead would be an interesting project...

I happen to signup for a 30 day free trial for a service using my credit card. Yet I always put it aside to try it until my first payment was sent. I guess shorter trials will force the user to actually try it quickly and see if it fits for his requirements.