Since the first day we opened up Aircall’s beta, we’ve offered a free trial. Indeed, Aircall provides phone number and an app to manage phone calls collaboratively, so prospects want to check the quality of calls and test the collaboration features before jumping into a fully-fledged plan.
Fair enough. Free trial is a common practice in the SaaS industry. Our free trial allows you to get a number of your choice, in any country in the world and offers you a 20-minute credit to make calls.
One detail still: numbers and call minutes cost us something. And not just a few cents.
Apart from that, our free trial was really going smoothly and helping us convert customers.
But on November 10th, there’s been a small hitch.
We were spotted by a blogger who boasts a modest 600,000 Youtube followers and 1,200,000 likes on Facebook.
He posted a very eloquent video – although it was in Arabic, we understood it pretty clearly 🙂 – explaining how to hack our free trial, creating a fake, temporary email address in order to get the activation link, choosing a random company name, picking a number (strangely enough he chose a number in Australia) and taking advantage of 20 minutes free calls from our app.
So in the morning of November 10th, our Slack feed-room started looking like this:
…and our analytics showed a small peak in visits 🙂
Yep, 54,000 signup attempts in one day.
Luckily and by sheer coincidence, we had updated our free trial registration process a few days earlier, to prevent direct access when using a non-corporate email like gmail or Mailinator. We were actually fed up with our 5 to 10 daily signups with generic email addresses that never converted to actual customers. This change resulted in having 99.8% of the 54,000 signups fail to access the trial – saving us a few thousand dollars.
Upon my request, the blogger withdrew his video and apologized for the consequences – anyway, his advice proved useless for 99,8% of this followers 🙂
At this point, we stepped back a little and reflected on our free trial. Here’s our take on it.
Do’s & Don’ts of a free trial for SaaS products
I’d like to share my analysis on 4 free trial issues we reflected upon. This is NOT an exhaustive white book on free trials, but simply conclusions from facts we observed. You may not agree of course, and I’d love to get your views about these!
Should you even do a free trial if it costs you something?
A web expert told me once that he’d never accept to offer a free trial that would cost something – literally – to the startup (apart from very limited storage/server costs). Instead, he’d go for a video or a « fake » trial. But as non-obvious as it seems, I disagree with his assertion that a costly free trial is a no-go.
But as non-obvious as it seems, I disagree with his assertion that a costly free trial is a no-go. The main argument against a costly free trial is that you can’t scale it. Imagine you reach the hockey stick inflection point or you get on TechCrunch’s front page, the excessive number of signups would burst your trial budget.
- My free trial can’t kill me cost-wise. In any case, my company’s credit card has a few thousand dollars credit limit, so worst case scenario, I spend them. But it’s actually quite unlikely because for it would mean a really fast burst (like our November 10th event) and my homepage would go down before the bank blocks our account.
- If ever I reached the hockey stick inflection point and saw my trial costs surge, well it would be a rich man’s problem : I could decide to let go because I believe this traffic is worthy or temporarily shut my website down to think deeper. In any case, reaching this new, faster growth rate does not depend on being able to cope once with a cost problem.
- In the end, it goes down to ROI, with only 2 questions
- how much does your free trial generate in cash vs. how much does it cost you?
- how does this ratio compare to alternatives (such as a demo video)?
About how much does your free trial generates in cash vs. how much it costs you
In our case, 10% of our trial customers were converting to a paid plan (it grew up to 30% now that we have a credit card step, but I’ll come back to that later). One month of a paid plan is around 40 times the cost of each trial user.
Without going into complicated customer Long Term Value estimates, the figures speak for themselves. I have a powerful tool to convert customers.
This leads to the 2nd question:
How does it compare to alternatives?
This a something we haven’t tested yet. The problem is that alternatives to a free trial are expensive to implement, thus it’s not so easy to A/B test them. But my gut feeling with our product is: people want to see a real number actually working, they want to make a real call before subscribing to a paying plan.
Freemium model or full version?
We started defining our free trial with a simple idea in mind : it should enable ALL features, or at least show them to the user, in order to to show him the full potential of the product.
But there’s a second rule – learned on the job: your free trial should be built on validation points. In other words, it must answer this simple question:
What does the customer need to validate before subscribing for good?
It seems obvious, but we didn’t build our free trial this way at first.
With Aircall, we knew that a potential customer would want to check whether the phone numbers are up and running and if calls actually get through with a good quality. Period. Our customers assume that the 24×7 level of service or the message recording feature work – these are not validation points.
We started offering 100 minutes of free calls in our trial. Why? We thought that potential customers needed to make inbound and outbound calls to each and every country they were targeting, but they don’t… WE make these tests – our users don’t. They just make one quick call, check they can hear and be heard and that’s it.
Our current 20 minutes are still probably an overkill 🙂
Should you ask for the credit card information beforehand?
Of course, I heard the traditional debate: asking for the credit card information in the registration process cuts your signup rate by X – 10 usually – but at the same time, it improves your conversion rate – by 10 again :).
Some specialists would recommend not using a credit card as a filter to improve lead quality [see Lincoln Murphy’s take on this] and I do agree as a general rule. But in the end, various options can be tested and measured, and it probably depends on the type of SaaS business you’re running.
We decided to check it by ourselves, and here are our figures:
Adding the credit card turned out to be more profitable for us. The real value driver is that people UNDERSTAND you request a credit card to give them a phone number and free calls, that’s why our onboarding conversion dropped by only 50% (I’ll come back to this in details later).
So for us, all in all, we still get eight times more paying customers, with fewer costs and less support. Admittedly, having more people trying the product and potentially talking about it to their friend has a long term benefit, hard to quantify.
I’ve tried to step back from these numbers and draw conclusions. Before deciding for yourselves, you should consider 3 outside factors:
Before deciding for yourselves, you should consider 3 outside factors:
- How does the value of your product materialize?
- What’s your main acquisition channel?
- How much does your trial cost you?
1. How does the value of your product materialize?
The value of a SaaS either materializes immediately or with time. In our case, our product delivers an immediate value: making and receiving calls. A SaaS that delivers you an immediately, perceivable gain makes you think (unconsciously):
« If I make the effort of giving away my credit card information, at least I’ll get this immediate benefit.”
On the contrary, productivity tools such as Slack or Front, for instance, take time to deliver their value. Therefore, these types of SaaS typically offer a free plan without asking for the customers’ credit card.
2. What’s your main acquisition channel?
You wouldn’t guess how much your acquisition channels matter in this choice. Let’s go straigt to the point, this is my take on it:
If your main acquisition channel relies on sales, you have the opportunity to build an intimate relationship with our leads before they enter the trial, so they can trust you will protect their information. And even if they don’t accept to put their credit card, we’ve got their email and phone number so we can reach out to them and reassure them.
But if your main acquisition channel is marketing, and more specifically paid marketing such as Adwords, your leads don’t have this trust factor, so you should avoid asking for your customer’s credit card information right away as much as possible.
3. How much does your trial cost you?
Of course, if you trial has a cost, asking for a credit makes sense. You’ll have fewer leads, but they will be more qualified. It should not be the only deciding argument, in my opinion, growth matters more than costs for a startup, but it should definitely be considered.
Should you restrict access to non-generic emails (assuming you’re a B2B SaaS)?
It very much depends on your business, but in our case: absolutely! Rejecting personal emails was key for keeping a clean funnel and making sure we only onboard motivated customers.
It did overlap with the credit card request, but in the end, only two of our paying customers signed up with their personal gmail account – and it did generate us support later when they decided they wanted to migrate their profile to their professional email.
So, should you block them or just filter for further investigation?
Recruiting software company The Resumator blocks generic email addresses. You simply can’t sign up using your personal email. We did it softer, we filtered them so that we could manually check them and grant access to selected addresses that we felt could be interesting.
In the end, offering a free trial is indeed time – and money – consuming, and it forced us to make some changes in our onboarding process, but it is key to our customers’ persuasion and satisfaction, and we will definitely go on with it.
If you have other advice and lessons learned concerning free trials, feel free to share them with the community in the “comment” section. And if you want to try Aircall for free (using your work email and your credit card :)), we’ll gladly welcome you!