To SaaS or Not To SaaS: That is the Question

[Revised, rewritten, and replacing a post from yesterday]

One question we encounter with our Information and Media customers is whether they should buy MarkLogic Server and build an application on top of it, or use a SaaS offering (which may or may not be based on MarkLogic) and effectively rent the use of an application to meet their online publishing needs.

The primary arguments in favor of the rent (SaaS) approach are:

• You get up and running faster because you’re renting the use of an existing application
• You have lower up-front fees because you need neither to build your application nor buy the hardware/software platform on which to run it
• You can focus on what matters because you are liberated from the nitty-gritty of building and deploying production systems

The primary arguments in favor of the build approach are:

• You create a unique offering which you can use to differentiate from your competition
• Your costs are potentially lower over the mid-term (SaaS’s relatively high annual payments reverse the initial savings over a few years; if you don’t believe me, remember that Wall Street values a dollar of SaaS revenue at about 2-3x a dollar of perpetual revenue)
• You create a strategic platform on which you build future applications, reducing the marginal cost of experimentation and new product development

To me, SaaS is not a religious issue; it’s a practical one.

While we typically sell our software on a perpetual license basis, we nevertheless are a big user of SaaS solutions at Mark Logic. We happily use Salesforce and somewhat less happily use Netsuite. I was also a champion of bringing Salesforce into Business Objects, where we became one of their earliest, large enterprise customers. (As I told IT at the time: if you won’t treat me as a customer, then I’ll go find someone who will.)

Turning back to the question of publishers and SaaS, like most questions in business, the answer should derive from strategy.

• If you are trying to compete solely on the basis of your proprietary content, then you should consider a “rent” strategy.
• If you are trying to compete on the basis of mixing content and its delivery mechanism, then should consider a “buy” strategy.
• If you are in between, then you’ll need to figure out where you are on the continuum and what you’re willing to trade for what.

As I always say, there are two things that money can’t buy: love and competitive advantage. Applied here, if you can rent a solution then your competitor down the street can rent it, too, and no amount of application configuration is going to result in competitive advantage (or disadvantage) for either of you.

What does this mean? It means that SaaS is great for what Geoffrey Moore calls “context” and rotten for what he calls “core.” Excerpt from the referred page:

Core – See Core/context analysis
Any activity which creates sustainable differentiation in the target market resulting in premium prices or increased volume. Core management seeks to dramatically outperform all competitors within the domain of core.

Context – See Core/context analysis
Any activity which does not differentiate the company from the customers’ viewpoint in the target market. Context management seeks to meet (but not exceed) appropriate accepted standards in as productive a manner as possible.

That’s why we happily use Salesforce and Netsuite at Mark Logic — we aren’t trying to differentiate on the basis of our accounts receiveable or pipeline management systems. (We are trying to differentiate on technology, market focus, and services excellence.)

So, for publishers

• The more your basis of competition is ownership of a proprietary content set, the more delivery becomes context, and the more you should consider SaaS

• The more your basis of competition is (1) uniting your content with other content, (2) delivering content in unique in-context ways, and (3) rapid innovation in online product development, the more delivery is core, and the more you should build custom applications (i.e., new information products) on a standardized platform.