Correlating trialware expiration with date of purchase

It makes intuitive sense that customers who complete the trial* version of your software are more likely to purchase it, but it cannot be so easily deduced that the converse is true, i.e. that non-finishers are much less likely to buy.

A potential customer may use your software twice and purchase it immediately, or purchase it much later without ever finishing the trial. However, it would depend on the nature of your software, especially for games and one-use software like importers and converters.

Have you, as a software developer, seen any clear correlation between:

1. Date of first use or download,
2. Date of trial expiry (whether 30 days or 20 uses), which can be inferred from the date of first use if not measured.
3. Date of purchase

One should also consider the % of buyers who purchased during and after the trial period. The conclusion drawn from this data would influence the optimal length of a trial period, the extent of crippling following expiry and the preferred method of trial limitation.

* fully featured for limited time or uses.

FYI, my question was closed on StackOverflow.com, so hopefully it is more suited to this audience.

I'm not sure this question has a general answer.

Yes, I've seen a correlation. No, I don't think it applies to other software.

It all depends on what your software does, how often people use it, now much of a pain point it solves.

Quick answer to your two questions:

1) Trial expiry: As a rule of thumb, start with 30 days trial, and then devise of a series of test that help you determine if you should increase of decrease the period.

2) I personally believe that your trial should only limit the stuff that's actually unique to your product, while still enabling anything that's available for free elsewhere. This allows people to stay with your product, and limits them from turning to a potential competitor.

Longer answer: Our company is a B2B business. We provide banner analytics for online ad campaigns. Our first users usually become champions of the software within their business. They are commonly not the ones that pull out the credit card. We offer a basic free version, and 30 day trials on any upgrades. This has worked well so far, but if anything we'd extend the period if trial. Our customers works in a process that is a) long (8-12 weeks), and b) prone to missing deadlines (becomes longer). The idea is for the customer to get a taste of the product before buying, right? If we were to limit the trial further it's likely we create more problem than incitement to purchase.

Unfortunately I'm sure you'll have to find this out for your own product.

Many ASP members have run detailed studies on your question. The most important thing to understand is that there is no general optimal length of a trial period, no optimal extent of crippling, and no optimal trial period. Everything depends on your application and target audience.

There are, however, a number of important things that have been revealed by these studies. Buyers can be classified into a number of groups:

1. Those who buy without a trial
2. Those who buy within a day or two after installing the trial
3. Those who buy at some point during the trial
4. Those who buy when the trial ends (are forced to buy)
5. Those who never buy (they don't like your product or they got what they needed from the trial)

Most of the studies I have seen are for business software- so that's what I am discussing now. A well designed web site with good ad copy should be able to convert about 40% - 70% of the potential buyers into paying customers without a trial.

The majority of buyers who download and install a trial buy within a couple of days.

The next largest group of people are those who are forced to buy by the expiration of the trial. (Note that the nature of some products makes time limitations impossible, so a "trial" for these types of products is typically only a demo- it is so crippled the result can not be used.) Finally, the least number of sales comes from people in the middle of the trial period.

The conversion rate of people who download into buyers varries widely. It can range from 1 in several thousand to over 70%. It all depends on the product, the buyers, and your trial.

For my software, I allow the user to add a certain number of records and once this limit is reached then they can't add any more records. This way there is no trial period, they can still use the software unrestricted, and they can come back and use the software any time after they install it and it will work.

I have read an article (and I can't remember where) that quoted a percentage of users who downloaded software and either never installed it, or installed it and ran it once, or installed it and never ran it at all. If the user runs it once and there is a 30 day trial and they come back to it 6 weeks later to evaluate it then the trial has expired. Instead of stopping them using the software because the trial period has ended, it is probably better to get them to contact you and extend the trial.

There are lots of scenario's unfortunately and no silver bullet.