Lately there is been quite a bit of bashing of the Freemium model. The last example is given by Ruben Gamez of Bidsketch. While he makes some good points, his rhetoric seems incomplete, if not flawed, to me.
I am very glad that he’s making more money now than before, but chances are that something in his Freemium model was terribly wrong and he was literally cannibalizing his business with it. What I would have liked to see in his article is not just a pie chart showing how many free vs paid users he had. That’s pretty useless data.
The breakdown in active users vs non active and the cost of each sector in total and by user would have been greatly useful for starter. The delta in drop off after the removal of the free plan. The growth rate not just for the first month after removal but for the months afte that. Then his pie chart would have been much more meaningful.
Let me explain:
If 99% of his users are Free and 99% of that doesn’t use their account, his cost for them is very much close to $0.
They might not be the right customers. They either will never turn into paid customers or it’s the wrong timing for them.
They might like the product but they have no use for it right now. Chances are they will recommend it to other people. I love DabbleDb, I have the free account, I never use it and don’t have a use for it. Recommend it to dozens of people. My wife is a paying customer of theirs.
Ruben’s cost for them in this case is the disk space required to store their login credentials, totally negligible.
If your product is good, a portion of those free users will spread the word and you can chalk up that expense to marketing. I have multiple free accounts with the usual suspects, basecamp, wufoo, chargify, etc. that I don’t use but I like. It’s for me to explore and evaluate their product. While I don’t buy personally I do recommend them to many people that end up buying or I might buy with the corporate account. (hence I don’t result as a conversion)
(Possible Scenario) If only 1% of the free users are active and 0.8% of all free user accounts are converting. It really means that 80% of your free users account are converting. Which one is it, I don’t know. Not enough data was shared in Ruben’s article to make that determination.
His says:
If I stayed on this path, I’d soon have thousands of free users to support.
Why? You don’t have to support free users. You do if you want to. Don’t feel pressured to, though.
Here is an idea:
- Make it clear at signup that only paid customers gets free support.
- Free customer gets paid per incident support. Your time is valuable.
Now for the opposite case:
If 10% or more of that 99% users were actively using the product. By actively I mean they use it at least as much as the paying customers. Beside that your conversion would still be pretty good, 0.8% of 99% with 10% active users is ~8% conversion, enviable. Chances are you either priced the premium plan wrong or your free plan is too good. It might be in fact so good that you could charge for it.
The right thing to do at this point would have been to a/b test charging for the free plan let’s say $2-$7/mo and see how many people signed up for it. You don’t even have to code the logic for it. Just test their behavior, then if people are purchasing you code it to charge. Grandfather those accounts or charge them later.
Still chances are that you needed a shittier free plan. Coming up with a free plan that works it’s not easy and it require a certain knowledge of human psyche and good analytic data.
If 40% or more of the free accounts were active you can be sure that your free account was too good and users didn’t need anything more than that. That was a plan you could have charged for.
Btw, there are many ways to make a free plan that still makes you money or that at least it doesn’t costs you anything.
How about: Ad-sponsored free plan.
For the first 2-4 weeks let them experience the site without ads. Then use transitional ads very much like wired, cnn, msnbc, monster.com. (click here to continue type) Don’t want them, feel free to upgrade. People will either drop off, upgrade, or painfully continue and generate revenue. Still these users are squatters and don’t get free support. No need to be a dick to them but make it clear that you’re busy supporting paying customers. If you want to be evil, alternate about 5-10 interactions without ads with just as many with ads. The constant alternating of showing ads/slowing down with free flowing is a really painful thing to experience and clinically proven to be much more irritating than continuous slowing down.
When you show ads all the time, people get used to it and get accustomed to them. They become white noise and the users block them off. If you alternate they can’t get accustomed to it. Beside making them more irritating they will become also more visible. If you don’t expect it, you’ll notice it more.
Promote to use
Make them tweet a predetermined phrase of praise to use the product every once in a while a-la macheist. If they won’t give you money, let them bring you users that will.
DabbleDb used to offer a free plan where all the data was public, if you can’t live with that, signup.
How do you curb the fakers and abusers? Tie their account to their twitter/facebook/disqus account. Let them login through a pre-existing account like facebook.
Imagine you’re this user, and you don’t upgrade what are you going to do? Are you really going through the trouble and pain of creating another facebook account just to have another free account. If you do, again it’s a clear sign that the free offering is too good and you could charge for it. Reduce the free plan even more.
If any free product is too good then by all means, don’t offer it.
If I were bidsketch, my free plan might look like this:
Proposals 1/month, 1 client, 1 user. No embed images, no custom domain, yes export to PDF, maybe templates and themes, no brandable proposals.
Lastly one more thing to consider is the churn rate. What’s the delta in churn rate after removing the free plan. If with the 8x registration growth you experience a similar churn rate you’re monetizing those free user for just a month or 2. You might still be leaving money on the table by not having that marketing working for you.
I have so many questions:
- was the removal done as a part of an a/b test? if not could the 8x increase due to something else? a positive review perhaps.
- Has the growth rate kept going at 8x vs previous month or is it just the new plateau (before x signup/months now 8x signup per month)?
- Has the growth rate decreased?
and many more.
The strength of a carefully thought out Freemium plans is not in the make money in early stages or the business model. Its strength is in the getting the word out and people to talk about your product.
First of all: It’s a marketing strategy, folks! Not a business plan.
A marketing strategy that will change over time. Once the word is out and you’re profitable, and having a free plan doesn’t really bring any more benefits: go ahead and remove it. It has done its job. It’s time to apply a different marketing strategy.
The one thing that I heartily agree with Ruben is to not blindly copy other people’s strategies. What worked for them at a moment in time and space, might not work for you or even be reproducible. Every niche and period in time needs the appropriate strategy.
If that wasn’t true, why don’t you go ahead and try to follow Bill Gates steps, exactly like he did. Go to IBM right now and try to licence them a DOS-1.0 like product. Why not? It worked for him, right?
Times have changed and continue to change at ever increasing pace. Strategies from 1 year or even 6 months ago might not work anymore.
Make your own, chances are you’ll understand it better and you’ll own it.
I have to go back and finally finish reading The Design of Everyday Things. I’ve had this book for a very long time but I can’t never see myself to read it. I really don’t like the typesetting on it. I guess I’ll force myself. 
