Public investors love recurring revenue businesses. The average Software-as-a-Service (SaaS) company is trading at 4.8x 2011 revenue, a huge premium to the 2-3x revenue multiples for traditional license businesses. The advantage of SaaS businesses (both for public investors and for the managers who run them) is that they are highly predictable.
At the beginning of each year, most SaaS businesses already know that they will meet 50-80% of plan from existing customers alone. This would be like starting a new semester knowing that you already have “A’s” in over half of your classes! Not bad!
That doesn’t mean that SaaS companies don’t work hard, of course. It’s just that revenue is delayed from the original sales effort (i.e the marketing or sales that occur in 2011 will often have more of an impact on the company’s 2012 financial performance). This delay (while positive for predictability) is why SaaS businesses often consume more capital than traditional license models (which better match the cost of sales with the cash received from customers).
But this got me thinking, if SaaS is more predictable than license models, is there anything MORE predictable than SaaS?
The answer is Freemium.
In a Freemium model, marketing dollars aren’t spent acquiring paying customers, they are spent first acquiring free users, who will (over time) convert to paid users. If you assume that the paid users are converting to a subscription product (like Evernote or Dropbox) then in a given month you will receive cash from three different groups:
1) Premium users who converted in prior months who continue to pay you
2) New users you acquired in the current month who instantly converted to a premium subscription
3) Old users you acquired in each of the previous months who finally converted to premium plan for the first time
The remarkable thing that I’ve noticed is how long this conversion cycle can be for the third bucket. Many freemium companies are seeing paid conversions even 2-3 years after the user initially signed up!
The conversion of old cohorts of acquired users actually make a Freemium business more predictable than a SaaS business. To illustrate this point, imagine three software businesses selling software with each of the three different business models (License, SaaS, and Freemium) during 2010. Now imagine what happens to 2011 revenue if on Jan 1, 2011 all three businesses fire everyone in the sales and marketing departments but continue to support existing users.
1) License Model - since sales and marketing are gone, the company will have $0 of new bookings in 2011. The company (if it’s lucky) might collect 15-18% maintenance fees from existing customers. Predictability Score: D
2) SaaS Model - the company won’t have any new bookings, but since existing customers continue to be supported the only decrease in revenue will come from churn. Assuming 15% churn, 2011 revenues will be 90%+ of 2010 run-rate revenue (remember that churn doesn’t happen immediately). Predictability Score: B+
3) Freemium Model - Not only will the company retain 90%+ of 2010 run-rate revenue, it will also add new bookings from previously acquired cohorts. This actually means that 2011 revenues should be greater than 2010 without the company spending a dime on marketing or sales. Predictability Score: A+
While Freemium is more predictable than SaaS, it doesn’t mean there aren’t problems with the Freemium model. Freemium businesses have a lot of waste since a company pays money to acquire free users, 90%-95% of which will never convert. These users can also be costly to support so you may actually lose money on your free users. Your marketing dollars are also spent even further ahead of cash inflows so you’ll likely need to raise even more money than if you had picked a different model.
So pick your model wisely. And if you can get a Freemium business to scale, sit back and watch the revenue grow. It will be a beautiful thing!