I hate freeloaders. And I hate giving out free trials.
My hypothesis is that SaaS’s giving free trial just have either:
- A weak offer
- Weak targeting and/or
- A weak landing page
And instead of lowering your price, you should add more value so that you can charge more / reinvest the money to make it better, and so on.
So why test out free trials?
Many people ask for free trials. From hackernews to Reddit, people bawl their eyes out about how cheap they are and how they have to try something out before they give me their money.
Also seems like ‘free trial’ is standard SaaS practice. I shouldn’t write off something that’s standard practice, even if I have extreme prejudice against it.
Experimental setup
This is a far from perfect experiment because many variables have changed. I’ve changed the following material variables:
- Added free trial option.
- $19/mo → $9/mo now, with feature gating. Previously, the only difference between plans are number of articles.
- Prior 3-tier plans are $19/mo (4 articles, unlimited AI-edits), $39/mo (30 articles, unlimited AI-edits), $499/mo (lots of articles). All plans are 4K words max.
- Now, the 3-tier plans are $9/mo (4 articles, 1.5k words max, 50 AI-edits), $39/mo (30 articles, 4K words max, 375 AI-edits), $699/mo (unlimited articles, 4K words max, unlimited AI-edits).
- Removal of 30d money back guarantee. Prior, I had 30d money back guarantee because if you’re serious about using the product and make an earnest effort to try it out, you can always just ask for a refund. Now, since you’ve got a free trial to try out some of the features, I will no longer accept refunds since they already know what they’re getting into.
Because there are some variable changes, I will only consider free trials as effective if it drastically increases what I want to measure.
Acceleration
As you’ll see below, there’s not been a ton of traffic to my website the past 8 months. To accelerate this experiment, I’ll do some guerilla + low-cost marketing in order to drive more traffic to get better stats. Otherwise, I’ll need to wait another 8 months. More on those sub-experiments to drive traffic, later.
What I want to measure
I just want to measure conversion rate.
- CR is conversion rate.
- T is traffic / sessions given by google analytics, up to and including Feb 23.
- S is subscription.
Don’t care about churn nor LTV. I just want subscriptions because as I’ve mentioned above, subscriptions are now non-refundable.
I need this to be drastically different to consider it a success. The reason for this is I don’t have a ton of subscriptions thus far to yield a very strong confidence in my current conversion rate. Thus, a more drastic conversion rate difference (probably 2-3X) is required for me to acknowledge free trials as something useful.
Preliminary user stats (pre-free trial)
There’s 3.3K users since July 1 to Feb 23, inclusive:
I exported all of my past Stripe payments, where each row is indicative of a separate payment. I want to grab the number of unique subscriptions, so I need to count in all of my Stripe payments, how many unique emails (unique emails in column Q) there are that’s paid me:
Which yielded 33 unique subscriptions. (Above formula needs to be changed to Q2:Q84 — going too high makes it div by 0 error, and we start at Q2 since we don’t want to include the header as a unique email).
I screenshot this in case you think it’s cap, because the numbers are a little too perfect.
Current conversion rate: 1%.
Goal conversion rate to validate free trials: 2-5%.
Caveat
There was one person that came to my website through insiderapps.com, and they started a free trial and paid $39/mo. I didn’t want anyone to sign up yet TBH because I wanted to snapshot my stats before pushing the news that it now has a free trial into the public. I also didn’t want anybody to sign up because the feature-gating and trials are buggy.
So buggy I accidentally deleted all of my customers’ past articles, unrecoverable.
Good thing Wraith Scribe lets you back up all your articles via email though so the data shouldn’t be 100% lost.
Anyway, I won’t count this new user as a free-trial conversion because:
- In my logs, I didn’t see them use any credits / generate any free-trial articles.
- Seems like they just paid $39/mo without doing anything in the free-trial.