SmoothSpan Blog

UnreasonableMen wrote recently with some concern and some great information about the profitability of SaaS companies relative to conventional software companies.  The article is a must read for anyone interested in SaaS, and most of the hard facts may be summarized with the central diagram from a MacKenzie report:

It’s easy to see why the analysis seems alarming.  SaaS vendors only turn in a 13% EBITDA versus the whopping 31% from all traditional software vendors.  The article then focuses on very high SG&A expenses and wonders a bit whether investor interest in SaaS is a bubble waiting to burst with no real returns as witnessed by the low EBITDA.

I don’t think SaaS is a bubble for a whole list of reasons.  One of the responses to the blog post was the comment that SaaS has real demand, where the Internet bubble was characterized by selling a lot of products that nobody wanted.  That’s an excellent obsrvation.  In addition, take a close look at the chart above and you might also conclude there’s really no point in doing software of any kind unless you can reach a scale over $1B–EBITDA is just as low for all intents on traditional software vendors that are small.  This has been true for as long as I can remember in the software world.   Until a company reaches signficant scale, and that scale is usually associated with a monopoly on some segment, software companies don’t make money.  There’s no point getting excited about the problem now–those horses left the barn long ago!

These economies of scale, by the way, are one of the major drivers (along with low cost debt due to low interest rates) behind the LBO and acquisition crazes for software companies. 

When I look at these numbers, I see some basic facts every software company CEO needs to keep in mind.  First, there are economies of scale, so size matters.  Companies that aren’t growing at significant rates need to think about mergers and acquisitions.  Second, fueling that growth is a top priority, hence the emphasis on big spends for marketing and sales.  If you’re lucky enough to have a product and space where you can generate big growth numbers you’d be crazy not to work as hard as you can to grow as quickly as possible. 

Third, COGS is going to be higher because of service costs.  Very large traditional companies are bouyed by huge maintenance streams that are essentially free from a COGS perspective.  Smaller traditional and SaaS companies have a huge service component.  SaaS buries it in the subscription fee, but provides data center and IT services, while traditional companies break it out as part of their professional services business.  The services business is an incredibly important source of calories for today’s software companies.  If there is any doubt about the impact of service costs, take a look at the EBITDA’s for traditional service companies like Accenture and these margins won’t seem so bad.

Another take away is that the desire to spend for growth in SaaS companies, which have a day in the sun where you can actually spend productively for growth, has been charged directly against their ability to do R&D.  This short shrift is compounded because SaaS companies have to build more than traditional software companies because of the nature of SaaS.  Traditional enterprise software has no need of multi-tenancy or a whole host of other things they take for granted but SaaS companies wind up building from scratch.  I think this is a much bigger concern for SaaS because it makes them capital intensive to create and grow, and it leaves them less budget on which to innovate, making them vulnerable to competitors who reinvent the wheel slightly better.

There is a discussion in the blog post as well of a Service Delivery Platform to help ameliorate some of the R&D costs.  I agree a platform is going to be essential to the overall health of the SaaS market, and a true SDP will do well there.  The economic pressures alone will drive its adoption if the SDP can deliver the right stuff.

The last observation I’ll make is that any customer for SaaS ought to view these numbers as an indication that SaaS companies are not taking unfair advantage of them and probably represent a good value for their money.