The Freemium business model has been on the rise lately (but it’s not new), largely due to the ease with which scalable cloud based solutions can support it. CRM and related platforms/services are no exception. Whether it’s a 30 trial, or a no cost perpetual set of reduced functionality, it’s being tried over and over again as a means to lure subscribers that could convert. It’s a risky business if you don’t know what you’re doing. Some reverse-engineered analysis of the product BaseCamp (and other 37signals offerings) suggest that (at best) the paid subscriptions account for 1% of the user base. That may not sound like much, but it can still be very profitable because the mouth of the funnel can be huge.
For those exploring the possibilities of this model, there is fairly simply math available that can help you understand where you break even, etc.. For instance, you can calculate the Contribution Margin (revenue per unit/period less variable costs per unit/period) which is the basis for a number of useful calculations. One fairly important thing to know is how many units of paid subscription are required for you to break even. If you take your fixed costs for the period (say month) and divide it by your contribution margin for the same period, you will end up with the number of units that need to be paid per month to cover your costs (fixed and variable). So, is that all you need to know when you’re running a Freemium business model?
What Behavior Pattern(s) Lead to Paid Conversion?
Decades and decades of trial and analysis have built some pretty solid analytical frameworks (some from before even the Baby Boomers, oh my!). Here’s the good news, they still require creativity to get results but they also rely heavily on left-brain thinking. While the formulas are tested, you still need to figure out what to put into them. That’s where the fun always starts – and sometimes the head banging.
Here’s the premise. We want to understand the lifecycles of various entrants to our Freemium user base. There isn’t one lifecycle, there are numerous lifecycles we need to understand. For instance:
- Which users will never become paid subscribers no matter what we do
- Which users will stop using the free service no matter what we do
- Which users can be convinced to upgrade to the free service
- Which users will upgrade with no additional investment on our part.
Sounds pretty simple doesn’t it? One of the problems for a new business is that you won’t have this data on day one, or on day two. However, you’ll never have it if you don’t build the capability to collect the data into your platform. It’ll take some time to see things develop. But here are some things you will likely notice right off the bat:
- You will see see a quick and sharp fall off among all new Freemium customers after the initial burst of activity.
- The ones that are not going to stay will tend to drop off even quicker
- The users that are likely to stay and become paid, at some point (you figure it out) will ramp up activity and their behavior will change to begin using more features.
- The remaining extended users will eventually drop off as Recency becomes extended
“Oh boy! He said Recency again! I’ll bet the next thing out of his mouth will be either “RFM” or managing disengagement.”
Don’t get all upset. I’m just restating what Jim Novo said to me recently. (Here’s a “do follow” link to a broader discussion of analyzing freemiums at his blog). He’s been working on some projects analyzing this very type of business, so if you’re thinking about this, or doing it, and have questions, I’d reach out to him for help. He definitely knows what to do!
If you can spot patterns in the drop off’s (disengagement, Ha!) that you can identity, you should be able to retain a percentage of them by delivering the right promotion, discount or whatever. Heck, they’re already here, how much does it cost to keep them here and try to make them paid? Answer, it’s nominal at worst. It gets back to know what to say, know when to say it, know who to say it to and know where to say it.
Now Comes the Fun Part
Now that you have some customers actually converting to paid, it’s time to grab a segment of them and take a look back at their specific activity and behavior during their free period. These patterns should be predictive of a transition to paid subscriber. These patterns will most likely appear as low Recency with high Frequency (see RFM!). Another thing you will likely spot is that subscribers who convert to paid will, at some predictable point, begin using a wider variety of features on a consistent basis. All you need to do is figure which of these behaviors turns into a paid conversion. You may also want to look closely at whether these behaviors come with a higher variable cost – such as a heavy use of file storage, maybe.
So, the question I will leave you with is this. “Do you have the tools in place to segment paid subscribers and identify their behaviors & patterns before they became paid?” Google Analytics will only get you so far.
