Panos Moutafis, Ph.D.
Case Study: Surprising Audience Insights
This is a short explainer video about our camera analytics using real event data. Our service processes vast amounts of information and insights; this is just the tip of the iceberg.
Enjoy and contact us if you have any questions! :)
Hi there and welcome to our video. Today we will be focusing on the “Camera Analytics” tab. To spice up the presentation, we pulled data from a real deployment and I will be using my story-telling voice.
For context, all the information you see on the screen was collected from a single camera which was installed in a workshop-style event.
First, I would like to draw your attention to the “Head Count” graph to the right. It tracks the number of people over time inside the room. On this specific deployment, the same speaker
offered the same presentation three times, back-to-back.
The first peak in the graph right here corresponds to the first group, the second peak to the second group, and so forth. In summary, three different groups of people came in and out and they all listened to the same presentation offered by the same speaker.
The interface also shows the breakdown by sex and age group. But the most interesting insights were obtained from the positive sentiment graph.
Let me zoom and show you what I mean. The positive sentiment for the first group was high, it got closer to the baseline for the second group and it dropped below average for the last group of people.
Our hypothesis was that maybe the speaker got progressively tired and could not connect with the audience as effectively.
Conference organizers invest a lot of money to bring the best speakers to their event so some pushback is expected when presenting these results. For example, someone might say, maybe there was a difference in the audience composition which resulted in the positive sentiment to drop.
The good news is that our analytics software offers a demographic breakdown over time.
So let’s take a look.
As you can see, the composition between the three groups was remarkably similar for the sex and the age groups. This reinforced our hypothesis that the speaker most likely lost some of their excitement (which totally makes sense) and maybe a few breaks between sessions would be very very helpful.
Our software also displays the positive sentiment level for each demographic group. On the left you can see the positive sentiment for the two sexes and on the right for each of the age groups.
Advanced users might appreciate the ability to adjust the smoothing applied to the graphs and controlling the statistical sample. Increasing the statistical sample will remove data points where there are not enough people in that specific group. This increases the confidence in the results.
Before we conclude the video, I would like to emphasize that all the graphs are updated live which offers the opportunity for immediate course correction. Each graph may also be exported for sharing within and between teams.
In summary, on this video, we covered most of the functionality provided by our camera analytics tab. Moreover, we used real data to highlight the insights and value provided by this amazing technology. But there is more. Much more.
So stay tuned for the next video!