“This time it is different” is a notoriously bad way to think … but, it is also obvious that one kind of business and another kind of business are not appropriately evaluated by the same metric and standards and one of the things about SaaS businesses is that it is in some ways a new business model with which we have as yet somewhat limited experience.
Clearly, selling hardware and software are very different businesses because one has to manufacture each unit of the hardware while much of the expense of the software is upfront. I say much since enterprise level software … as opposed to something shrink wrapped … can often include some enhancement as a part of each sale. Back when I was first selling my ERP software I could make some money on the hardware as well since the medium sized companies I was selling to had limited IT departments and one needed something mini-computer-like to support many companies. Now, the same number of users can be comfortably supported with better performance on a PC running Linux.
The genius of SaaS is that while one forgoes the big paycheck at time of sale, each sale turns into substantial recurring revenue for a significant period of time. Moreover, it is attractive to the customer because it eliminates the big up front cash outlay to purchase the license and often eliminates the cost and hazards of managing one’s own hosting (SaaS licensing can be used for on prem installations, but most often is not).
SaaS does not eliminate the need for customer-specific modifications, but it provides a huge incentive for the software company to think about better design. Pre-SaaS, many software providers were either “over the transom”, i.e., the customer was responsible for all customization or, if the provider did the customization, that was often done at minimal cost to create a unique implementation for that one customer unless the feature was one that was a good candidate for inclusion in core product. With SaaS, there is a strong motivation for the software provider to think in terms of a unified implementation with “switches” to invoke different behaviors for different users so that a single unified product was providing as much of the functionality as possible for all customers.
Which said, done right the SaaS business model enables extremely rapid company growth … as we have seen for some of the companies we discuss here. As we have seen, this strains our valuation models. I would like to suggest that this is exactly because of the rate of growth. We have seen that some traditional valuation metrics are invalidated by different business models, e.g., the Amazon keep investing the potential profits in growth so that P/E remains low despite enormous success. The inherent problem of an extremely high growth company is that a huge part of the “valuation” that the market provides is based on expectation, not performance. So, when those expectations are met, then the expectations continue to grow, but if the expectations are not met, even by a little, there can be a huge … technically disproportionate … contraction of expectations and thus of price.
At times, we have talked about the silliness of quarterly earnings reports and their impact, especially since analyst’s expectations are often based on little real knowledge. But, with these companies we would almost wish for monthly updates so that we could more rapidly adjust our expectations. And yet, even then, much of the reaction would be lacking in understanding of what was actually happening with the company.