There has been a great deal of discussion over the past few years about whether unified communications (UC) is actually being adopted among enterprises, or whether all the talk about UC is just noise generated by vendor marketing departments. I have seen myriad surveys conducted by various organizations showing that the respondents had either:
* Deployed UC technology
* Were in the process of deploying UC technology
* Planned to deploy UC technology over some near-term planning horizon
* Had no plans to deploy UC technology
* Etc.
These surveys are interesting but they are, of course, just a sample of the overall market, and it is difficult to extrapolate from surveys because the samples often have a selection bias. Furthermore, surveys are just snapshots that don't show clear trends unless the same sample is resurveyed periodically.
Last summer, after a conversation with friends, I thought that there must be a way to get some non-vendor-biased authoritative data that would tell us what was really happening. Fast forward 6 months and we have the answer.
My Methodology
I have been capturing data on actual deployments from a "UC crawler" that I have constructed with some partners. Like a search engine crawler, this process traverses the public Internet, looking for evidence of UC deployments at the edge of company networks. Note that this doesn't provide a complete picture, as some organizations choose not to allow UC to traverse their network edge, or their traffic traverses the Internet over VPN tunnels that block visibility to the traffic; and obviously we can't see behind the enterprise firewall.
We did a quick study of around 140 enterprises that we know have deployed UC technology and found that our methodology was able to detect around 100 of these. Therefore, we can estimate that we only have visibility of about 70% of the total number of actual deployments. Extrapolating from that data, the number of actual deployments could be up to 43% larger than the numbers shown below.
The crawler is capturing this data from over 355,000 (no, not a typo) companies around the world, so we think that we have most of the potential deployers of UC in the data set; particularly the large organizations such as publicly listed companies and non-commercial entities, of which there are nearly 38,000. The remaining entities on the list are small private companies.
UC Deployment Data
This data is for enterprise deployments of UC, and not the usage of UC by individuals via a "peer-to-peer" technology such as Skype; note however that we can detect the usage of cloud-based services such as Office 365. For our purposes, "UC" traffic is defined by the protocols used, specifically SIP and/or XMPP although we are aware of (at least) two vendors who use both, so that introduces some double counting. Our data can be broadly categorized into 3 groups:
* Federation & Outside User--These are deployers of SIP technology both for inter-company federation as well as for roaming users.
* XMPP deployers--Broadly similar to the SIP group
* Other SIP deployers--We are still trying to get a clear definition of what this means, but our best guess is SIP Trunking.
For anyone who wonders if UC is a fad or an actual technology transition, the absolute figures are interesting in themselves:
(Note that the totals don't quite add up as some organizations have deployed more than one type of technology.)
As you might expect, the propensity to deploy UC is strongly correlated with organizational size. Of a list of nearly 7,000 publicly traded companies from around the world, we see an overall 14.4% deployment rate, with US companies leading at 19.4% apparent deployment.
Deployment Growth Rates
We have been capturing the bulk of the data since the New Year and we re-run the survey around every 2 weeks. Although we don't have a statistically significant sample yet on growth, the rate of growth that we are seeing is quite surprising. The SIP Federation and Outside User deployment increase from February 21 to March 22 is as follows:
(Note that the XMPP and "Other SIP" categories are too new to comment upon)
Let me stress that this is a single data point, so my college stats professors are certainly rolling their eyes at this point. Although 4.3% looks like a high number (~66% compounded annually), it is building on a low base number so, even if that rate of increase holds up over time, we won't reach 50% deployment for 5 years. However, the technology adoption curve shows us that deployment rates go non-linear once a technology starts to become mainstream.
Summary
Hopefully I have gone some way to provide solid answers on UC deployment. Per above, the growth data is a work in progress, and our methodology continues to be refined. However, the absolute deployments and positive deployment trend present pretty clear evidence that UC may have "crossed the chasm".
Those who are interested in gaining access to this data will find further information on my website: http://www.ucinsights.com/.
