The ongoing discussion on various aspects of Software as a Service prompted me to think about some differences which may not be apparent to investors.

The barriers to entry for on-demand application solution providers (SaaS) are low. All you need is a reasonable idea, some smart design and good coding to get going. After that, it’s down to customer intimacy, marketing (not always needed though) and effective service delivery (always essential). You don’t need to have any massive investments in underlying platforms - you can buy that as you grow. SaaS has low entry barriers, which means lots of competition and duplication, unless you build a trusted and valued market offer, brand and customer franchise. Investment returns can be spectacular, but easy market entry means they may not be durable (i.e. multi-decade durable).

Sub-application platform services (PaaS or Platform as a Service) hosting multiple SaaS solutions, and providing everything below the application functionality, e.g. user administration, billing, security, operations continuity, load management, servers, comms , etc etc etc) require serious upfront money and a very high level of skill in building and operating them. Also, instead of marketing to the masses, or even business in general, you’re marketing to the web-savvy segment of the IT industry - actually quite easy, if you’ve got a good solution and reputation. And your service is equally valid for bespoke and single tenant applications as well (replacing in-house servers). Increasingly, this is a game for major infrastructure players or those with the ability and capital to become ones. Returns are likely to be lower but should be more durable.

These two variants of the SaaS opportunity are very different strategy, investment, development and operational propositions. That’s not to say that good application providers can’t make the grade to smart platform operators. Or that good platform providers can’t move into applications. Just be wary.