One of the recurring topics at the recent Enterprise Connect conference was interoperability, or the lack thereof. Jim Burton and Fred Knight moderated a packed session on interoperability and asked the audience if the vendors were doing a good enough job--deafening silence.
The topic came up repeatedly at the conference as a point of frustration.
Most of the vendors insist they are committed to standards and interoperability. Here are some sample clips from vendor web sites:
* Cisco involvement in the open community is demonstrated by its activities in open standards, open source, and open extensibility, to develop the potential of the Human Network.
* Polycom is and always has been committed to open standards and interoperability.
* Microsoft runs the UC Open Interoperability Program (UCOIP) to qualify third-party vendors' enterprise telephony infrastructure. The program helps customers have a seamless experience with Microsoft Exchange and Lync for setup, support, use, and end-to-end call quality.
* Siemens Enterprise: This means that our customers benefit from knowing they can integrate best-of-breed technologies that complement our own portfolio, that they're protected from technology lock-in and can learn from the experiences of other customers who are using similar solutions.
* Nortel and Avaya share a commitment to open standards rather than proprietary architectures and have been leaders in SIP innovation.
* IBM continues its commitment to working with leading telephony, audio and video providers so you can take advantage of the most advanced collaboration capabilities available.
With all this commitment, one would think that interoperability would be fairly easy. But it isn't and it's actually getting worse. It reminds me of my college days where we stopped calling beers by their names and used their can slogans instead: "choicest hops," "purest water," "Artesian wells"--"fully committed to open standards"--rhubarb, rhubarb, rhubarb.
Voice PSTN and email are the primary UC interop success stories, and these are not exactly recent triumphs (even pre-UC). Other than PSTN protocols, phone systems have very little interoperability. Try putting a call on hold or park on system A and retrieving it on system B, or sharing busy light status between different systems even when both are on the same private network. Voice dialing and email are powerful tools of communications because of ubiquitous interoperability. Though, new interop issues lurk with the upcoming SIP transition to IPv6.
Here's the painful fact about commitment to standards--it's only a starting point ,and effectively it's just rhetoric without additional effort of alignment. Simply stated, standards are not enough. SIP, for example, has a high degree of variation in implementation. Vendors and carriers are still manually testing/approving supported partnerships because the standards leave too much to the imagination. Of the thousands of SIP providers, major vendors only support a handful or two, as the certification process is too cumbersome and expensive. As a result, standards-compliant systems can be totally incompatible with one another.
UC changes the interop story from a comedy of errors to a tragedy. Communications is busting-out of the confines of voice. New devices, applications, and capabilities are creating far wider interoperability challenges: presence, phone status, conferencing, collaboration, video, mobile phones, etc., all expected to seamlessly work together. Vendor component oper-ability (inter-what?) is a challenge in itself. An interesting character in this story is the UCIF (UC Interoperability Forum), but it isn't yet clear if it is to be the protagonist or martyr.
Almost a year ago, the UCIF formed to tackle this mess. The goal was not to create new standards, but to create specifications for interoperability within the standards. This model, an industry consortium, is fairly well proven when the vendors agree and work toward resolution. But not all the UC vendors agree. The UCIF rapidly got 38 members to join-up, but Cisco, Avaya, and Skype are not among them. Cisco and Avaya together dominate the enterprise equipment market and Skype's probably got more UC traffic than the rest combined.
That hasn't stopped the UCIF, at least in goals. I asked Bernard Aboba, the UCIF chair about this back in November 2010. He said:
UCIF has a broad mission which ranges from certification of devices which might fit in the palm of your hand all the way up to room-scale systems. This spans a wide range of industry niches, each with their own ecosystems. So to evaluate potential effectiveness, you have to look at each effort individually. At the moment, UCIF has three Task Groups, in the areas of Devices (USB Audio and USB Webcam) and video (H.264 profile), and in each one we have convinced a broad cross-section of the industry to participate. More and more companies are signing up, and at least one Task Group has 40+ members. So at least for those groups, the problem is probably less about "critical mass" than achieving consensus.
You don't need everyone to create an interoperability solution, but you do need everyone, particularly market leaders, to implement it. I ask various folks at non member companies why they have not joined, and typically get a response like "what has the UCIF done [in its first nine months of existence besides recruit 38 vendor company members and create its structure and priorities with numerous task groups and workgroups]?"
While users are clamoring for interop, the vendors are measuring their commitment to standards, while the one industry group with potential to improve the situation languishes without support from the market leaders. A no-win scenario: participation requires visible progress, which is effectively impossible without participation. Instead, we have all these vendors working independently on the problem of working together.
How important is interop? Consider presence: the pitch goes something like "only call people when people are available", "a huge productivity gain", "reduce trips to voice mail", "why wait until office hours if you are both working?"--rhubarb, rhubarb, rhubarb. All these claims are valid, but enterprise presence is generally limited to enterprise networks and our work flows are not. Very few enterprise presence systems interconnect. Federation is a weak solution. It means that if both companies have the same or supported solution and upper management in both firms agree that it might be good to share presence, then some users can. Several prerequisites so that some users can interconnect with some colleagues. If that isn't sufficient, most brands can connect to public IM networks (AIM, Yahoo, Google)--but only some networks. For example, MSN’s network is closed. The net-net is most enterprise presence solutions are islands.
The next version of Microsoft Word and the rest of the Office 15 suite is expected to offer integration of IM/presence with Facebook. Neither solution is particularly open, other than their open disregard for Google. The move further blurs the distinction between personal and business communications and networks. But instead of connecting two closed networks, wouldn't open interfaces in in each platform, and others, create a powerful ecosystem of partnerships and tools? This raise-all-ships strategy seems obvious, yet remains only rhetorical.
Meanwhile, Skype is laughing its derriere off, with its closed UC club serving over 600 million users--with about 30 million online daily. Skype membership and usage continues to skyrocket, and its planned IPO valuation is rumored at $1 billion. Turns out this UC stuff really works: nearly all my Skype calls go through because I check presence first. Sometimes I even IM before voice or video calling. Skype users made 102.5 billion minutes of calls in 2009, equal to about 20% of total international calling. Skype believes 37% of its users conduct business over Skype while the rest of us leave each other voice mail. Skype doesn’t offer rich integration with any enterprise UC solution (Skype's SIP Connect does not support rich UC or presence, though theoretically this is under development).
Lack of UC interoperability is a real problem that is getting worse, not better. I know many vendors have interop labs and work hard to certify various relationships, but this model wasn't effective before and the challenge is getting bigger. The problem is scalability. Consider all the permutations of webcams, smartphones, and tablets expected to hit the market in the next 18 months; at best hope for a partial list of devices with limited integration.
The productivity benefits of unified communications are powerful and transformational. Current solutions are driving and enabling work processes in ways unimaginable just a few years ago. However, lack of interoperability leaves considerable potential on the table. Real commitment to interoperability would increase capabilities for all users as well as the value of all UC products and services.
For additional information: This post offers some examples of interop progress. To learn more about the UCIF, see this recently published white paper which I co-authored.
Dave Michels, Principal of Verge1, contributes frequently and blogs at PinDropSoup.com.
