Patrick Moran of New Relic had a great piece on TC the other day, talking about revenue in SaaS companies. The most interesting part to me was this:

Bad: Churn can kill you, or at least your market cap

If you are building a SaaS business, churn is your enemy. Most public SaaS companies report their monthly churn rate, either as a percentage of revenue or actual customers gained/lost. These rates depend on the type of business – 2% monthly churn is in the “tolerable range” according to many experts.

Patrick is right. Churn is terrible for a business model that relies on long-term relationships. If you're a car dealership, churn isn't as big of a deal, because purchases are discrete events and only occur about once every 7 years. So if you screw something up, you have a lot of time to correct the problem (or at least enjoy many years where you can still produce a lot of profit).

In SaaS companies, one way of thinking about churn is in the context of switching costs and sunk costs. One example of this is an API, or rather what APIs represent -- money and time spent on integrations. APIs are a form of competitive moat that protect a business thanks to switching and sunk costs -- if customers spend time and money integrating with your system, they're not going to switch as easily.

Note: I'm not encouraging lock in as a good business practice, but as an investor, it's worth considering the switching costs at a company (both switching to and from). Lock in is actually something we strongly avoid at Cloudspace -- for one example, by not locking clients into fixed development schedules, we earn our business every week, and that ensures we have the right incentives. Then again, we're less Software-as-a-Service, and more like Skills-as-a-Service (or maybe Brains-as-a-Service), but I digress. If you're a good company and you treat customers right, it can buy you a lower rate of churn and lock in isn't quite an issue.

This all being said, inverting the problem is interesting. Churn at your own company is bad, no doubt. But, if you're competing against an existing SaaS company, you want to increase your competitor's churn. One strategy for doing this is to build easy-to-use importers that import a customer's data from their system into yours. By reducing the cost of switching to (or at least testing out) your service, you can grease the adoption funnel. As far as I know, music streaming services (Spoitfy, Rdio, etc) have never done this. I don't understand why. Eventually it would produce a bit of an arms race, but that's the nature of competition.

To be fair, a company like SugarCRM vs SalesForce has more complex data structures, so this isn't quite as easy as a SaaS with simple data structures like playlists, but there are still strategies for dealing with this. And of course, there are strategies to defend against this, too ;)

If you want to learn more about APIs, check out the O'Reilly API strategy book or talk to my friends at Mashery, who are outstandingly good with these things.