Ruisrock

Bringing Emotion Into Customer Insight:
RUISROCK VIDEO ANALYTICS

As an event organizer would you like to know what your visitors are really experiencing during your event instead of relying on the rare answers to survey questions? What if you could get real-time insight into people’s specific emotions at any given time, stage, queue, shop, band or even during a song? When Ruisrock, the most popular summer music festival in Finland, asked for CGI`s help to get more information on their customer experience they got more than anyone could have ever imagined. Welcome to the era of neural networks and facial analytics.

Until the present day, customers’ engagement with brands has been three steps behind. Why? Because the first problem with traditional feedback is that it`s just feedback—opinions based on past experiences and expressions of current states of the mind. The second problem is that you don`t actually get much feedback at all, and the feedback you do get usually comes from two extremes: things are either very good or incredibly bad. There are simply too many variables and opinions rising from whatever happened and they might not have anything to do with the event itself. The third problem is that traditional feedback is very general. At a large event like Ruisrock, based on the feedback, you will have no idea which part of the event created the best experiences and which did not.

For these reasons Ruisrock`s partner, Mellakka Helsinki, contacted CGI. They wanted to know whether it was possible to do better than this. And we answered their call.

Our creative solution

At CGI, we knew that we could provide Ruisrock much more value. We could significantly improve their visitor experience by providing emotional insight inside their festival gates. We pitched Ruisrock with the solution on how facial analytics and its underlying technology of neural networks could reveal their visitors’ emotions in real-time and across locations.

By placing IP cameras to cover each of the 32 entry gates and the primary VIP area and, additionally, equipping 10 – 12 groups of volunteer workers with GoPro HERO 5 cameras, we were able to get over 310 million data points of information from analyzing the video taken from the visitors’ faces.

By using our custom neural networks, we got realistic estimates on festival goers demographics (age & gender) while capturing other facial analytic information such as emotions (micro expressions) and visible expressions (did they look happy or not) all in an exact place and time. We also used the GoPros to record audiences during select performances and for the first time ever, the event organizer could know the visitors’ exact emotions on any given time, stage, queue, event, shop, band or even during a specific song. It is hard to imagine a better way to improve your customers experience!

The implementation

The planning started in early 2017 and the team delivered the project in only three months of service design workshops, planning and programming.

During the workshops and meetings, CGI, Ruisrock and Mellakka Helsinki worked together to define how to best capture the emotions and demographics of festival goers. This data would meet Ruisrocks goals to understand customer segmentation and provide sentiment analysis. We realized we would get the most festival coverage and best promote Ruisrock the best by also utilizing GoPro cameras in addition to having gate cameras. Once the plan was in place, the technical planning and programming was done by the CGI team.

CGI provided all the hardware and once the system had been built, tested it prior to setting it up on the festival grounds. The data flow was automated, the data visualization was configured and all was ready to go for the event.

The team was able to set up and test the onsite installation within one day. People were assigned specific duties, such as administrating the volunteer GoPro setup and memory card swaps, technical operations, data visualization and marketing. During the event, we had the operations team monitoring all of the equipment so that we could react to the multitude of issues that arise at large scale events, such as this.

Technologies utilized

We implemented a multi-tier platform to mitigate the risks we would incur with network availability. We used edge devices for static cameras, an on-premise server for GoPro data processing and a cloud solution for all data collection.

This platform allowed for on premise scalability, reliability and distributed processing while benefiting from cloud data storage, advanced analytics and processing during off hours when the network was available.

We used edge processing on the gate and VIP cameras and stored the data when the network was not available. Our on-premise server processed all of our GoPro content. Once the data was processed and stored on the server, the data was transferred to the cloud.

All of the onsite data visualization was done off of the server so that we did not have to rely on network availability. Advanced analytics and associated data visualization were done in the cloud.

The solution is GDPR compliant.

The results

In addition to the few thousand written surveys Ruisrock collects every year, they now have more than 310 million data points. Talk about face value! The data from the visitors was already utilized during the event as we identified happiest customer segments and happiest ages for both men and women. In the future Ruisrock will use this data to further improve the event and its customer satisfaction by providing services that best boost the current emotion and demographics.

Ruisrock

310 million data points

Ruisrock

1.6 million faces analyzed

Ruisrock

Over 300k GoPro location events

Ruisrock

On average, each festivalgoer was analyzed 2.73 times

Ruisrock

Ruisrock got comprehensive insight into each visitor segment, as the traditional surveys cover a much smaller group of people.

Ruisrock

The insight confirmed that the happiness was measured highest near Sateenkaari stage, right next to the entrance. That’s also where Ruisrock had placed a giant unicorn and other visual effects. This finding spurs Ruisrock to put even more effort in the entrance experience in the coming years.

Ruisrock

Emotions, be it positive or negative, form a strong tie between people, and enforce the memories of the event they shared together. The large scale of emotions recorded at Ruisrock enforces the theory and gives a human touch into analysing the data, and a new perspective into planning the festival.

Gender distribution

Based on user survey (%)

Ruisrock

What our data showed (%)

Ruisrock
women
men
survey by Research Insight Finland

Age distribution

Based on user survey (%)

Ruisrock

What our data showed (%)

Ruisrock
under 20
20 - 24
25 - 30
31 - 40
over 40
survey by Research Insight Finland

Data is being used to understand and identify correlations of festival events and locations and their sentiment impact over time. Examples include: stage and vendor locations during different performances, non-performance or vendor areas over time and the main gates upon festival goer arrivals. All of which provides insight into how people related to the festival, across events and throughout their stay.

”We`ve had succesful years previously, but we wanted to know what we could do even better. Of course we have always done the traditional surveys, but never before have we had a chance to study peoples’ emotional states during the festival. The biggest realization of this survey was that, what a large scale of emotions can be experienced during the festival and from that we can draw conclusions on how to affect people’s memories of Ruisrock by offering right services at a right time!”

Annakaisa Anttila
producer, Ruisrock
Ruisrock
Ruisrock

Team

Annakaisa Anttila

Mellakka Helsinki

Communications office

Tuomas Välimaa, Mellakka Helsinki

Images

Saara Autere, Vessi Hämäläinen, Jussi Virkkumaa

Ruisrock

Team

Aarni Alasaarela, Riz Haque, Eero af Heurlin, Teea Hurme-Rintala, Krista Jouhtimäki, Mikael Karppinen, Vesa Kuoppala, Päivikki Kyykkä, Edward Mauser, Anna Savisaari, Juuso Viljanen, Julius Wang