Managing across the voice-data divide

One of the most interesting issues to emerge from the IP telephony research we’ve done at Nemertes Research is the challenge of getting voice and data teams to work together. Part of the problem is political: IP telephony often has been positioned as a way to eliminate all those “useless” voice specialists from within IT organizations.

Leave aside for a second the fact that anyone with such a mindset obviously fails to understand the sophistication and critical importance of voice communications. The real issue is that such an attitude virtually guarantees the failure of your IP telephony implementation. You just told a key group of people that the successful implementation of a project will involve the end of their jobs. Hmmm. What do you think the chances are of that project being implemented successfully?

Moreover, voice specialists often are as well educated as data personnel, have spent more years in the field, are held to higher standards – and are paid less. There are historical reasons for this (which tend to illustrate the general principle that you mess with the free market at your own risk – but that’s another topic).

The net result is that data folks often have a chip on their shoulders about the inherent superiority of “their” technology, while voice people carry a burden of fear and resentment that hinders them from maximum productivity.

To fix this, first, managers should reassure the voice specialists that they won’t lose their jobs at the end of the IP telephony deployment. They should do their best to address glaring salary or other compensation inequalities (there’s often a 25% or more pay discrepancy between voice and data specialists).

If you haven’t procured IP telephony products, have the voice and data teams separately create selection criteria and evaluate the vendors separately, promising a final “bake-off” for both sides at the end that uses the combined selection criteria of each.

I bet that the selection criteria from the voice and data teams won’t overlap more than about 60%. You’ll have an enlightening session or two during which you’ll have to explain the definition of “mean opinion scores” and “call completion rate” to data guys, while expounding on the difference between packet loss, latency and jitter to voice experts. The process is worth it – you’ll be far more confident in your vendor selection, and your voice and data staff will have a greater appreciation for each other’s knowledge.

Once the selection is complete and you’re ready to move into deployment, have the voice folks benchmark call quality and user feature sets while the data gurus do the quality-of-service benchmark and infrastructure assessment.

Finally, once the deployment is complete, ask voice folks to manage users while data folks manage the infrastructure. Why? Frankly, voice people tend to have more experience and more patience with stupid end-user tricks. Patience is a rare and precious quality these days.