Why Tibbr Matters

Sorry, this will be a quick one so I don't have time to build in an Animal House reference. I've been following all of the "expert" commentary on Twitter about Tibco's Tibbr announcement. At first glance, Tibbr appears to be another entrant in a crowded space that includes Yammer, SocialText, Salesforce (sorry again, a quick one so I don't have time for what deserves to be a very long list, you get the idea).

Tibbr is not a me-too entrant in this space. Tibbr leverages Tibco's unique position to drive highly relevant information into an incredibly compelling new knowledge worker experience. The link between critical line of business data and the average knowledge worker in an enterprise has long been more tenuous and less effective than it should be. Tibco has an investment in enterprise application integration that might astound the uninitiated. So here is an example of how an internal microblog/activity stream might look in a Tibbr-enabled world:

"Joey's Diner has the best clams casino. I can't get enough of them."  — Get over yourself, no one cares what you're eating. And by the way, your self-indulgence is creating noise that is blocking critical signal.

"The Knack are back. Great concert last night at the Konocti Harbor Inn." See above, you get the idea.

"I'm working on a huge proposal for a key client, has anyone done a big deal with legal"   Great, now we're getting somewhere.

"Inventory levels of widgets are critically low in Kansas City store"  Holy mackerel, that's important. Thanks, Tibco, for making a massive investment in building integration to my inventory system.

"Employee Mary Smith has requested vacation time in excess of her available balance. Click here." — See above, you get the idea.

Now, it's not clear to me that Tibco needs to be in the world of building and selling software for knowledge workers. There is likely an ultimate separation between content and user experience. And all of the vendors fighting for a real estate in your enterprise knowledge worker's world probably need what Tibco is providing with Tibbr. A beautiful home needs plumbing, gas lines, etc., or you can't live in it. This a powerful step forward for enterprise social technologies and anything but a me-too.


The challenge lies with the activity sources

First off, I agree with you analysis. I think Tibco is in a much better position with their integration background to make this work. They have access to both messages posted by people, as well as messages posted by systems. Therein lies the barrier, though. If those systems aren't publishing anything interesting, the value will be limited. It's great that someone manually posts a message about inventory levels, but that's always going to be slower than an automated message from the place where the important event was first created/detected. I posted a little bit about this theme when Tibbr was first announced in Dec. 2009 (http://www.biske.com/blog/?p=734), and the premise still holds true. We need to design out applications to share events and activities that previously were never published outside of the internal processing world of that application. Until that changes, tools like Tibbr will be limited by a weak information pool.

Challenges redux

The challenges Todd highlights are very real and analogous to problems early on with Enterprise Logging, BI, as well as snmp and alert notifications. The capabilities of the supporting operational tools are only as good as the information they are able to intake and make sense out of. It took a while to convince development to throw standard exceptions that could be captured and reported by Tivoli, for example, and logged in a manner that was conducive to centralized logging and reporting. I can't count the number of business intelligence or decision support projects I've been on where the technology was great but the available data pool was lousy (or nonexistent).

Not just any event

I also second Todd's point. But, in addition you need quality event publishing, not just a vast stream of undifferentiated events. Systems need to publish what matters, and what is published needs to contain enough information for events that are meaningful for some people to make their way to those people and not get swamped with meaningless noise. This is exactly the lesson we should have learned from SNP (as Chris mentions). I have seen numerous systems that publish vast numbers of SNMP alerts where critical alarms are swamped with messages that everything is going fine.

I also second Todd's point!

Thanks for the feedback everyone. Great stuff. I'd love to get this conversation going because I think there are huge opportunities and huge challenges.

Setting up the integration is a huge and non-trivial start to business value. Tibco has a unique ability to access the data. That does not preclude the next task which is delivering the right information to the right person in the right context. I think this will ultimately become a business level analysis exercise that will leverage the technology now available to drive processes to stakeholders in the right form at the right time. Business and IT need to address this together or it will likely fail despite some cool emerging technology. One thing that back-end systems have access to that is very valuable is lots of meta data that can define potential audiences and be delivered as a payload without users needing to define the value. There are some fairly basic examples of defining where someone needs the information delivered to their activity stream, but there is a lot of work still be done to define value. Certainly, the individual worker should have the opportunity to define what is valuable to them and should leverage back-end security. Tibco brings a lot to this.

So where do we go from here? As you all point out, data does not equal value and too much data degrades value. Preferences, context, analytics, system payload can all help define value to an individual worker. That's easier said than done, but ultimately separating signal from noise is critical to the value of enterprise social.

Posted a followup blog to this...

I've posted some additional thoughts on my own blog on this topic: http://www.biske.com/blog/?p=811