Cisco CMX Analytics in use at CiscoLive! Orlando

Editor’s Note: This blog post was co-written by Dr. Brendan O’Brien, who is authoring the post, and Dr. James Little, technical lead for CMX Analytics in Cisco’s Wireless Networking Group.

This year’s CiscoLive! in Orlando with over 20000 visitors was  selected as an ideal opportunity to test and prove CMX Analytics over a large venue with high volume usage.

This year’s CiscoLive! was spread over two huge venues of the Orange County Convention Centre and the Peabody Hotel between June 23^rd-27^th. Over two and a half days, approximately 40,000 different devices were detected using CMX Analytics. (This figure is not uncommon any longer, assuming 50% of people had WiFi on and most of us had 2 or more devices)

The main purpose of the Wi-Fi network at CiscoLive! is to provide a high quality service to all  attendees, providing CMX Analytics with its toughest test yet. With a peak of around 15,000 connected devices on the opening day serviced by 750 access points and 1500 radios, the challenge was to see how well CMX could take the raw numbers and translate it into insights on the movements of the delegates.

Let me break down some of the reports for you:

In the table above, CMX Analytics shows:

• The second floor being busier than the 1^st, 3^rd or 4^th during lunchtime. as expected with the catering facilities located there.
• The World of Solutions was also busy over this period as it was a break between sessions.
• The keynote areas are expectedly quiet during this time.

Valuable insights can be provided to the venue operators, such as the busiest parts during the day, common movements, and dwell times. This gives the venue owner feedback on how to plan staffing levels, refreshment facilities and signage.  Furthermore for the Cisco network team, the analysis verifies whether the allocation of APs is in line with where people are spending their time. The value of having this information available to the venue operators during the conference means that adjustments to layout, staffing or service  can be made in a dynamic manner.

CMX Analytics also shows individual paths and, more importantly, aggregated behavior across the venue. A typical path a delegate followed between starting from the ground floor can be seen in the below graphic:

Focusing in on one floor, we see the same numbers repeated over larger time windows, showing that the mornings were the busiest in that part compared to the afternoon or evening.

Taking this a step further and using CMX Analytics to drill deeper into the World of Solutions: As you can see in the graph above, on the 26^th around the middle of the day, people are seen spending more time looking around during the lunchtime window, with the average dwell time reducing as the afternoon progresses. Clearly this shows that for any exhibitor the peak times for high quality of the engagement is during this ‘lunchtime’ window.

In summary, CMX Analytics on its first outing to CiscoLive! shows how the information captured on the attendees can provide valuable insights to assist with the event.

If you’re interested in location-based services and analytics, I highly recommend you take a look at two new whitepapers James and I published recently:

• A Technical Review of Cisco’s Wi Fi-Based Location Analytics
• Transforming the Mobile Experience with Cisco Wireless Location Services

For more information on CMX, visit www.cisco.com/go/cmx.