Why Education is Not Enough for Video AnalyticsBy: John Honovich, Published on Mar 28, 2010
While industry consensus of video analytic's disappointing market performance has been reached, one school of thought suggests the solution is more education on what video analytics can and cannot do. See a recent advocacy of this position from SDN citing IMS:
I attended an IMS Research press conference and was informed that 2009 was a tough year for video analytics (which wasn’t exactly surprising since 2009 was a tough year for everybody), but the most interesting part, I thought, was that in what they cited as the solution. One of the primary reasons that video analytics has been slow to adoption is due to a lack education of both end users and integrators about what the technology can and cannot do.
Examining guidelines from Cernium's Archerfish's limitations and the optimizations needed for Aimetis analytics shows the limitations in the pro-education position. Also, compare to our analysis of install requirements for 3VR's facial recognition. [If you have not read them, I strongly encourage you to do so before continuing on.]
Let's imagine that these companies and all analytic companies were frank and upfront about these limitations early in the sales process. Two things would happen:
- Customer satisfaction would improve significantly.
- Sales would drop significantly.
We can debate how much each would vary but it is clear that both would move in opposite directions. As prospects understood how hard it really was, lots of people would decline. The rest would either fit as is or be willing to make the expensive adjustments.
The status quo, being unclear or reticent about the technical barriers, helps get analytic companies in the door and into sales they would otherwise not be considered.
While it is doubtful that this would make the market grow, it would increase real world success (which would clearly be a positive to users). On the other hand, since most of these companies are venture-funded, I doubt they are excited to trade off revenue growth for smaller numbers of satisfied customers.
The mass market does not want to buy products with cumbersome restrictions. You would not buy an HDTV set that glitched 15 times a day and that started acting crazy when the sun shined on it or it rained hard outside. The same goes for video analytics.