UC: Big Promise, Modest Movement

The buzz around enterprise unified communications is loud, and getting more so as IT spending loosens. The problem is, in our experience and confirmed by our InformationWeek Analytics 2010 Unified Communications Survey of 406 business technology professionals, enterprise-wide UC programs that have a truly transformative impact on business processes are all too rare.

For example, videoconferencing has lately hogged the spotlight. But too often we see IT groups set up expensive video systems and walk away, with nary an hour of training or any plan to track whether employees even use the tool. From the CFO's perspective, consumer-class applications, such as Skype and Yahoo Messenger, seem to provide much the same benefit as enterprise-class systems, without all the hassle and expense. No wonder we're faced with frustration, misunderstandings, and elusive ROI.

Trust us: Your users know what's out there, and they know when their IT departments are hunkering down or worse, dismissing their needs. One wary survey respondent says of UC: "While this would cut down on wasteful e-mails, I suspect it would ramp up the number of distractions an employee has in any given day, and that it would consume a lot of the employee's time with little return."

When you start confusing collaboration with wasting time, you're seriously missing the point.

There's still time to turn this ship around. The market forecasters think it will happen, with Infonetics Research predicting the enterprise UC segment will grow from $256 million in 2010 to $398 million by 2014 in North America, according to a May study.

But it will take a few key commitments on IT's part, some easier to accomplish than others. First, build training into the budget--and we don't mean allowing for a help desk staffer to spend an hour lecturing employees. Then, leverage business intelligence principles to help drive UC adoption and speed ROI. Make very sure your network is ready for the load, and plan ahead for interoperability by keeping an eye on standards.

And finally, cede control to the business units. That one might be the toughest, but it's also the most important. As we discuss in depth in our InformationWeek Analytics UC report, available to subscribers at information week.com/analytics/uctoday, technical glitches are generally seen as the major cause of disappointing UC implementations. In reality, IT often brings problems on itself by ignoring the end-user experience and failing to have business leaders set the strategic direction for the project.

You Are Here

IT groups must understand where they are now so they can figure out how to get where they want to be. The "now" isn't exactly impressive for most companies, based on our survey. "We have done a casual evaluation and don't see enough benefit compared to the cost," says one. More than half of those not deploying UC see it as a lower priority compared with other projects, and 34% say they see no definitive business value. Ouch. Just 10% of those with UC plans have a non-IT C-level executive playing a key role in developing a strategic vision. Only 14% rate integrating UC with enterprise applications as a very important priority. The No. 1 and No. 2 keys to success? Voice over IP and unified messaging.

What is this, 1997? And why the dispirited outlook? Are IT teams facing insurmountable technical or interoperability hurdles, or perhaps nonexistent budgets?

Not in most cases. Of course, the concept of "virtual" everything has captured the minds and pocketbooks of IT, and to be fair it's much easier to show the cost savings of going from seven servers to one than it is to show how much you saved or earned because your client got to the right employee faster or a videoconference saved dozens of phone calls. The biggest challenge, however, is poor user adoption, our survey finds. Technology can help here; the aforementioned high-definition and desktop videoconferencing applications are impressive and can entice employees into at least trying the system. But there is no killer application that can singlehandedly pull a UC project out of the bog.

Speeds And Feeds

While getting employees to actually use UC applications is the single largest factor in a successful rollout and in achieving ROI, it's not all about soft skills and selling your services. Ensuring your infrastructure is sized to handle the traffic is also very important. In our experience, IT gets only one shot to hook users on a new application. If it fails out of the gate or provides suboptimal performance because of network problems, you're sunk.

We asked respondents about their top concerns regarding their networks' ability to provide appropriate quality of service to UC applications. Aging infrastructures with limited WAN bandwidth top the list. These concerns are valid, especially if you need to provide Power over Ethernet.

Video, the current UC darling, does require a significant effort to transport. Thus the relative dismissal of carrier WAN service-level agreements was very surprising; only 4% of respondents say this is a concern. To the other 96%, we'd tell you about a client we worked with several years ago that used AT&T to connect its bank branches. IT was assured that it would get bandwidth that could "burst" to a full T1, which seemed sufficient for UC. When VoIP was deployed, the company experienced very poor performance. It turns out that the carrier was offering a "burst" of traffic by dropping packets into a virtually bottomless queue somewhere in its cloud. Traffic got through all right--after a short 800- or 900-millisecond delay. AT&T assured us that it was within its SLA for bandwidth because it did in fact deliver the packets, instead of dropping them, so buyer beware.

And remember: Trust but verify. You can see drops and delays on your gear, but you can't see them on your WAN provider's equipment, especially if you're buying services like MPLS. It's critical to the success of UC that you negotiate, understand, and test your carriers' SLAs. You may wind up fighting battles that span weeks and even months to get UC applications like video to work properly. This wasted time and erosion of user goodwill kills your adoption and erases any opportunity for ROI.

Furthermore, IT must not consider UC purely a network initiative. "As we roll out UC, we are struggling with where support should fall," says an IT architect for a large manufacturing firm. "This is a new paradigm." Our firm will provide about 200 voice and video readiness analytics engagements this year, and about half the time, the "Windows" or "server" teams won't want to participate. They don't feel that UC will impact them, or that they're responsible for its success. They're dead wrong. If you're still engaging in turf wars between IT and telecom, we guarantee you're depriving your company of collaborative opportunities.

The Power Of UC Is In The "U"

Even with an end-user-focused approach, a unified IT team and a network that's ready to perform, there are still formidable integration hurdles.

"While UC hasn't been an overt strategy, components that could form the basis of a UC strategy are in place already," says Phillip Howey, business service manager at Nasco, a healthcare claims processor. "We have invested in collaboration and workflow tools as well. I would suggest that while we have made strides in the 'C' of UC, the 'U' is lagging." Howey is exactly right. To amplify the power of the U in UC, we need to refine connectivity between the communications silos that we have today. And not just within our borders but between organizations that may use different providers.

That's likely why our survey shows that companies prefer, by a 10-point margin, to patronize best-in-class UC vendors that support and promote integration, over single-vendor offerings that try and do everything under one banner. "The biggest barrier is the lack of standards and integration between the primary competitors," says an architect at a major U.S. automaker. "It's forcing you to make a bet on one company." We feel his pain. We also believe that companies that take a best-of-breed approach and hold their providers to account on standards will be better able to extract the full functionality from UC.

For the 45% who prefer to go with a single vendor, be aware that while the name on the invoices may be the same, interoperability still isn't guaranteed. That's because many UC vendors grow through acquisition. We won't mention any names, but we distinctly remember a certain major UC vendor shipping a "lite" version of its call-processing software and a "lite" version of its voice-mail software on a single box that was touted as the perfect small-office phone system.

Nice idea. The only problem was that the default transfer settings for the call processing software were incompatible with the default transfer settings for the voice-mail software. So when a phone forwarded to voice mail, the caller got a busy signal.

Of course, gluing communications silos together isn't a new idea. In most cases, multivendor UC environments have rudimentary integration for making calls. In the past, this integration was accomplished by linking systems via PRI and/or QSIG trunks. This got the job done but meant spending a good amount of money on gateways and line cards for two VoIP phone systems to call each other. Pretty stupid. Thankfully, Session Initiation Protocol has taken hold, and much of what the future of UC holds will depend on SIP becoming pervasive.

We think it's well on its way, but that does not mean that all integrations will now be seamless--again, standardization must remain a top priority. SIP still has many variants, many "extensions," and many quirks, such as DTMF relay mismatches, to be resolved.

Start Today

Still, you can start tying together the systems you have in place today even though there's still some work to be done to deliver seamless automated intercompany communication. Most vendors have embraced standards, certainly SIP. Even when there are compatibility issues, they're normally resolvable, typically without having to spend additional money, if the IT department is willing to collaborate with vendors to make it work.

UC integration requirements are, however, becoming broader in scope. Our old "PSTN" glue only handled voice. But now we need to handle video, IM and chat, and presence applications, not just voice communications. The power of MSN Messenger or Skype, for example, isn't in the desktop client, but in the directory.

Even more integration is needed if you're going to bring these consumer silos into the enterprise. Your salespeople, for example, want to be able to add any client on any network from their corporate IM or phone systems and reach them. If your top account has standardized on MSN Messenger and you use Microsoft Communicator, can you reconcile them?

The short answer is yes. This idea gives IT nightmares from a security perspective, which we think is misplaced, but it's true to the vision of UC and is what corporate VoIP and unified messaging must become.

We hear howls over the "misplaced" comment, but we stand by it. Every form of communication has an associated risk. Malicious hackers use the phone to call users and trick them into giving up passwords, but no one has suggested turning off phones. What you must do is educate your people, and then control as much of the communication environment as possible.

Lest you think we're awash in negativity, our survey does show positive developments. "UC has definitely transformed the way our business operates," says one respondent, adding that he cut telephony operating costs by more than half. The benefits are there. We just have to break out of our ruts.

First, emphasize how UC can make communications more effective and efficient. Collect business intelligence and create a strategic plan. We firmly believe that the technology exists to do almost anything you want. The missing link isn't the connection between applications or UC systems but the connection between what the technology has to offer and what employees expect and will use. Close this gap through training, targeted application deployment, reliable networks, and an approach that places the needs of the user at the top of the list.