« iPass CEO at Stanford’s ETL   |   Main   |   Thinking thru AttentionXML before EBIG SIG next week »

IBM surveys perception of On Demand mktng message

Last week I received an invitation to a survey offered by Meta Group on IBM’s OnDemand program. It didn’t seem to work the first time I tried it, but apparently I misread a question (bad survey design!).

It’s worth it though because it offers a glimpse into IBM’s marketing and insights that 1). they’re perceptive enough to suspect (and willing to spend money confirming) that their marketing is missing its target, and 2). that there’s conflict within IBM groups about which message should be most prominent. Take the image below, for example, which I captured out of their survey. This question presents seven perfectly reasonable interpretations of what IBM’s OnDemand initiative means:

-

Now, I can make a guess (until they share survey findings as promised) which option ends up being the most popular (#1), and the option IBM’s marketing honchos really want to see (#6) - but what’s truly interesting is the small, yet important, differences in what each interpretation means. Each bullet lists an empirically strong value proposition, which is clearly targeted to a particular functional group in the enterprise. So, this survey is a test to see if the right message resonates with the right functional area (the survey requests role and position). That’s very good marketing execution and perfectly valid.

However, if you take a step back, and look at how different the value propositions are within each of these, and how each value propositon really points to a different level of integration and reliance upon IBM - we see exactly how confused one can get about just what IBM is supposed to bring.

If IT is looking at IBM as an infrastructure provider, and the board is looking at this as a business model intiative (which I think is pretty hard to swallow), and the procurement guys see this as a sourcing model - things get pretty convoluted when you roll all these projects up, never mind the fact that each group is convinced that their idea of IBMOnDemand is correct.

We’re all guilty of relying on high concepts such as On Demand to sell products and servics - but in this case I actually think what IBM and HP and even Gartner are doing in this space is dangerous. IBM has its version of IBMOnDemand, which is both vague and big enough to safely cover all these attributes mentioned above. HP has “Adaptive Enterprise”, which, to my mind, is even less clearly articulated than IBM’s. Then we’ve got Gartner selling their “Real Time Enterprise” which, when I can parse their texts on the matter, seems to be an even higher conceptual model of this same vision of provisioned/billed-when-needed, autonomically-managed, integrated-via-components model.

This model itself is sound - it will be a cheaper, more efficient, more robust way to operate. The migration demanded by this vision is enormous though, as is the implicit flexibility demanded of these systems. I’m troubled by the monolithic approach of IBM and HP (check out this ITWorld webcast they sponsored with Gartner for an example - site registration required) in this space. Sun seems to get it more with their talk (well Jonathan’s is the only stuff worthwhile) of Solaris 10 containers for virtualization - that seems to offer a middle path, namely in migrating currently-running applications without modification.

What’s missing here, I feel, is the simple recognition that any of these utility schemes seem to involve significant engineering, along the rough equivalent of putting in a new electrical system in a house. Why would you want a proprietary wiring system? This sounds basic, but I’m not convinced that IBM and HP (and even Sun) get that these utility infrastructures need to be open standards and portable. Customers will sign up for service contracts, but it will be very difficult (as both HP and Sun have found) to get them to rearchitect their environment to work only with their products.

So the final question of the survey was to comment on both the good and bad of all this. I shared the following with Meta, which I’ll share with you here. I absolutely think that IBM’s amoeba-swallowing-approach is dangerous to what is a genuinely important next step in enterprise architecture, even though I do admire them for cooking up discrete and compelling value propostions for each of their constituencies. It’s a heck of a marketing campaign; it’s just not particularly helpful to the ecosystem.

Question: What is the single biggest thing IBM is doing right at this time?
My Answer: The words “on demand” capture what businesses need to be doing HOWEVER…. (see mistake - next question)

Question: What is the single biggest mistake IBM is making at this time?
My Answer: The IBMOnDemand (please note - it is one word, IBM’s version of this otherwise useful phrase of OnDemand) offering pollutes the concept of flexibility and modularity that should be inherent to any given generic OnDemand solution. The IBMOnDemand framework is all IBM [ed. note - not absolutely true, but it is WebSphere everywhere] - and thus very expensive to implement given the fact that we all sit here with heterogenous infrastructures that we cannot afford to rip out and replace with a beautiful greenfield solution delivered by cheerful and skilled IGS consultants. This mistake is potentially crippling, not just for IBM, but for the whole enterprise IT market. With this mistake they are in serious danger of destroying the positive mindset around SOA, components, utility, modular systems: component-based is supposed to mean the end of the need for (not, please note, the end of the desire for - there’s ample room for selling) large IGS-style projects. IBM pushes the world-altering nature of this under the rubric of “let IBMOnDemand be your new… business model” - surely you and IBM can see how crazy sounds for enterprises. It just doesn’t seem realizable (because, as we are reminded in the ads, IBMOnDemand is a perfect state) and it’s not something that you can meaningfully implement incrementally. Lastly IBM is on thin ice with this emphasis on providing business models to enterprises. I would never want IBM to drive my business model; they might help enable me to create my own, but buying IBMOnDemand is not buying, as you said in survey, “a way of designing business models and processes”. I’m all in favor of selling at the board level (I have a great tech analyst presentation here that counsels selling to “board level” - I’ll do up a post on it one of these days), but IBM’s marketing message is just too rich here.