But we’re looking now at and this came out for hackathon, we did with the data lab recently, they were looking at having cameras on exhibits, that measure the proximity of people and how long the people stay there. But the really nice thing about the test that was done at this hackathon, with data students, was that in this you know, a couple of days, they came up with a camera that, that with facial recognition, could track how people were feeling. So you get two ratings, you get how long people were spending there, what’s the engagement like and you know, were they happy about it, or were they really unhappy about it. If it’s a scary exhibit do they get scared?
Or if it’s, somebody’s gonna make you laugh, are they actually laughing or are they looking bored? So that’s, that’s I suppose that’s slightly for the future, but we’re working on that. We certainly can gather data, in terms of people’s feedback, but the opportunities that we haven’t exploited yet, need, need investment. So if we can put the technology and the infrastructure in place around the racecourse, then we can start gathering loads of financial data, about how people, where people are going. What they’re buying. How they’re moving around the racecourse. It’s not something that we’ve been able to do yet, but we’re very aware of, that other sort of stadiums and visitor attractions can do that.
We have people often contacting us, through messenger and Facebook, to ask maybe about our tours. What time they are at? Are they suitable for whatever groups of people? So we’re currently looking at developing a bot on Facebook, that will give people, I mean if people contact us at 3 a.m. in the morning there’s not going to be someone here to reply to them, so at the moment they just get a reply message to say these are our office hours, which isn’t maybe the best customer experience.
So we’re working on developing a bot that when people do get in touch, they will hopefully be able to give the visitor the answer that they need, through a series of questions, or if not put them in touch with the human who can help them. Because as great as data is I think there is always going to be need for a human somewhere and especially in tourism, human interaction is, is very important you know, people come to Edinburgh they want to meet locals, they want to have that human interaction. So, there’s lots of things that we can develop and lots of things that were looking at, but there is but it always going to be a human element.
If it gives people an answer to a question quicker then great, but if they want to talk to someone, there’s always going to be the human. So in terms of our future data development, the thing we’re really working on at the moment is, at what we have helpfully termed choice paralysis. So the festival is massive and it’s brilliant and if you’ve got the you know, the 2019 program, there’s three thousand eight hundred plus events to choose from. And actually we recognise that for a number of people, that amount of choice just leads to no choice. You just think, I’ll just go for beer instead, or you’ll go and you’ll find yourselves in the places you’ve always been.
So we want to, we want to explore, we’re working on a project with the audience agency at the moment around this. How we help audiences understand their tastes, their taste profile and help them stretch, the elasticity of the risks they might take. We’ve been talking about it more in terms of Spotify and how it feeds you familiarity and stretches the edges of you, rather than the sort of, the recommendation model that says, ‘you bought a lawnmower, would you like another lawnmower?’ So it’s, it’s about not recommend, is about not recommending. It’s about helping audiences find things, that they feel would be compatible to them.
We’re sort of thinking of it like a dating site, it’s not about finding the perfect show, it’s about finding shows, that you might enjoy to spend an evening with. And then we have to think about, how we’re going to incentivise feedback in that space and how we incentivise people to help us build and develop those taste profiles. And, and then we can layer in things like demographic and all those sorts of things and and then feed that information to shows, about how they might communicate and find their audiences. This is all about helping audiences and artists find one another. Not about saying here’s the top 10 shows you should go and see and it’s a massive challenge for us.
It’s quite, it’s quite infant and we’ve got a few little strands of work happening at the moment, it’s really exciting just to see how the the day science sector will come up, with interesting new ways. Erm, I’m expecting you know, would like to see some machine learning in there potentially, some AI human interface in there, that helps people understand what they like. And it’s less about good or bad , it’s more about did it, did it satisfy. What did it, was it meaningful in your engagement. And that’s very exciting for us and is a huge project over the next probably two to three years.