I made fun of Managed Objects new social-networking influenced CMDB interface product myCMDB when it was announced last month, critiquing both the kindergarten playground name of the software and the fact that something as prosaic as a configuration management database product was jumping on the seemingly ubiquitous social networking bandwagon in the wake of successes such as MySpace, Facebook, and LinkedIn, but Managed Objects are the ones laughing now. I finally got a chance to look at the web-based CMDB interface, which can tie in to Managed Objects CMDB360 product or a number of other CMDB backends, and it looks good.

The social networking aspect of the software only seems comical until you consider some of the challenges of populating and making use of the information that properly belongs in a CMDB. As has been pointed out elsewhere, the practical aspects of establishing and maintaining a CMDB are daunting. Dustin McNabb, Vice President of Marketing, lays out the case for using the social network approach very compellingly. "We have learned an awful lot in implementing CMDBs, particularly with our large users," says McNabb. What they learned was that while the information in the database was broadly useful and in demand at all levels of the organization, the burden of inputting and updating the information tended to fall on those who had implemented it originally: IT professionals, who ended up serving as very expensive data entry staff. While automated discovery systems were good at populating the basic objects in such systems, it developed that it was the relationships which were most valuable to the companies... and those could not always be automatically maintained. Consequently, certain "tribal" information would not always be included properly in the CMDB, diminishing its utility for decision support services.

The solution, Managed Objects CTO Kurt Westerfeld believed, was to distribute the system in order to reduce complexity and improve usability. Using the same principle as social networking sites such as Digg or Wikipedia, the burden of entering information could be pushed down to the lowest possible level of responsibility, who also happened to be the people most likely to have accurate knowledge about the systems they were working with every day. Conversely, making all the information from other staff equally available to those people directly and unmediated by a select group in the IT department was sure to pay dividends in its utility in their day to day job functions. Wrap it all up in a Web 2.0 inspired interface, and leverage the power of the crowd to overcome the problems. But will it? More after the jump.

There are certain situations where a few expert users are the best way to obtain accurate, meaningful data... you don't want to crowdsource your brain surgery, for example, or try to run a nuclear submarine by a show of hands. In other situations, a combination of a large number of normal users can produce results that far surpass those of a small, cloistered group of experts. The quandry in social networking today (and much of the debate over its efficacy is unwittingly centered in this as well) is to find which problems the second solution can be successfully applied to.

My feeling is that CMDB maintenance probably is an area where crowdsourcing could effectively produce results surpassing those offered by a relatively small number of maintainers from the IT department. But whether or not a sufficient body of users are going to be inspired to come together to produce those results is another question.

Managed Objects already faces an uphill fight to have this theory accepted by mainstream IT, and they are downplaying the social aspects of the software at this point, saying essentially that you don't have to use it that way if you don't want to, it's still a good interface into the CMDB. Which it is. But without the social networking aspects, it's an interface suffering the same challenges as all the others available out there, namely that few people use them and the maintenance of the information either languishes or becomes a terrible slog for staff with talents best applied elsewhere. Just because the interface is distributed doesn't mean the incentives are; managers outside of IT, with no clear stake in the system and with other priorities can't be relied upon to motivate staff to enter data in it. In other words, it doesn't solve the problem it sets out to unless it does become a social network, with compelling value to users at every level.

It's a bold move by Managed Objects and only time will tell if the approach will be successful. Unlike other sorts of software packages, which may be fairly judged at a demo or by features and performance, this sort of software depends primarily on the implementation and culture at the business which installs it. You can't buy a social network in a box; you can put software in place which may facilitate it, but you can't expect it to grow up magically around it; there have to be incentives in the system for people to willingly, even eagerly, use it. And this is the real challenge for the software, even if it can find favor in IT departments. Will it find favor with users?

That is the foremost question in my mind surrounding this concept: can something as nerdy and bland as configuration management truly draw in users the way Facebook or Digg can? McNabb believes it can; he thinks "it's a natural place to come to" for users such as database architects, support staff, and even administrative types, where they can hang out and exchange information on their favorite topics, which are presumably centered around database servers and network switching relationships. Maybe so; as I said, it depends greatly on the culture of the organization where the system is being installed.

But even if that's the case, how many social networks can a person reasonably belong to and contribute to actively enough to make them successful? I get fatigued between checking LinkedIn and Biznik. Throw in a couple e-mail accounts, IM, and more personal social activities, and does anyone really have anything left over for something like myCMDB? I don't know of any studies done on this, but while communication is good, and options for communication are good, it seems to me that at some point if you put in too many possible lines of communication in an organization there are diminishing returns; you lose the signal in the noise. People start tuning out. In an unrelated briefing the day before I talked to Managed Objects, a support vendor mentioned to me that the number of applications on the average desktop has exploded in recent years from 25 to 80. That's a massively increasing support burden, of course, but it strikes me that it's also a significant distraction to the user... to be familiar with and regularly use eighty applications? Something has to give.

My suspicion is that things like myCMDB are what will give, as interesting and carefully considered as the product is.

It seems likely that increasing numbers of vendors will attempt to use the lessons of Facebook to alter their products into social networking facilitators. But it may be that the very fact of all those "myX" products will be what keeps any of them from succeeding. When you have to check your inbox on myCMDB, myExchange, mySAP, etc, etc, to keep up with all the mundane events of your busy day, you're either going to spend all day skipping between them, or find ways not to use them at all. And social networking, unused, is unworkable.

What is needed is a sort of meta-social network, a product that can establish itself inside an enterprise and effectively filter up all the other data and narrow silos of interest that these other products attempt to monopolize, something to fulfill the promise that the corporate intranet once offered for one-stop shopping on the communications and information front. This, it strikes me, is one of the end-game benefits of Service Oriented Architecture, or could be in a company with enough forethought to build it. But it seems unlikely that any independent software vendor will be able to offer anything so grandiose, although no doubt many will try.

Without something of that nature, I fear that social networking in corporations will be entirely hit and miss, with some organizations finding success in a limited number of implementations of this sort, and others finding they they are mired in an excess of cutesy dashboards and inboxes, unpopulated and underutilized.