David Boyle argues that creative industries still rely on human judgment for core decisions, despite having access to data. He shows how BBC Earth and Snoop Dogg used audience segmentation effectively, but warns that data alone would have picked the wrong single in one music-industry case.
Chapters
So the year is 2015. Even slightly old fashioned creative industries like the music industry, book publishing are absolutely swimming in data now, whether it's on the production or the consumption or the distribution of their products, data is everywhere. It's pretty easy to query, pretty easy to put data science on top, pretty easy to visualize. It's available on most people's desktops. So you might assume that these businesses are becoming ever more data driven in the way they make decisions. It's definitely not the case, at least with the core decisions at the heart of those businesses. Which artists do I sign at a record label? Which authors or books do I sign for my publisher? The really important decisions have changed very little over the years. They're still made by gut instinct, or as I like to call it, skills and expertise. And it doesn't mean that that has to be the case. In the business I've worked in, you absolutely can introduce data science into those decisions. They can be embraced and welcomed and they can be used alongside and without treading on the toes of the creative vision. Sometimes I've seen that work more successfully than others, I should say. And I've learned a ton of lessons along the way, some of which I wanted to share with you today. So the first lesson is that you need to leave people in the business with insights from data science work that they can carry around in their heads. This is because in creative businesses in particular, most decisions, the vast majority of decisions are made in a pretty unplanned way. They're made in emails or phone calls, they're made in a meeting that was actually on a different topic, or they're made at the last minute. Because there weren't great planning processes in place, they couldn't have commissioned data science to help them with that decision. So you have to pre program people, you have to preload people with insights that they can carry around in their head so they can draw on them when the decision comes along. One way of doing that that's been really successful is through clustering or segmentation of data sets. It applies to lots of different data sets. Here's a great example, a company I'm really excited about called Afineo, we use them to understand our social media audience at BBC worldwide. In this case people who follow one of our brands, BBC Earth, in a particular country on Twitter. Really simple clustering of quite complex data science into something that even the most junior marketing manager working on the brand can really easily understand, who the different types of audience are, therefore, by the way, who they're not the different sizes of different audiences, the relationship between them. And they can monitor this over time. Dead simple. And most importantly, they can carry it around in their head so that when ad hoc opportunities for content or editorial or commercial partnerships come up, they can draw on this knowledge in their heads to make a better, quicker or smarter decision on the fly. Somebody else who would talk more eloquently about the need for this than I would if he could be with us today is this person he fell in love with, with data science. What we did in the music industry, now, Snoop Dogg might not describe it quite that way. He fell in love with the outputs of the data science. He fell in love with the segmentation and the opportunities that showed him to grow his audience. In particular, there was one segment we recommended from big data analyzed and clustered. One segment of consumers stood out as the next opportunity for him to grow his audience. At the time, they were called gadgeteers. Young people digitally engaged social media, always on their smartphones. Of course, today that's most people. That was a segment back in the day and it was the next opportunity for him. He leapt on it throughout the meeting, he kept interrupting, saying, we've got to get them gadgeteers. And you can see it in his actions. He took it to heart. The digital partnerships, the technology partnerships, his social media engagement. In the weeks and months afterwards, he totally went after that audience and he totally grew his audience in doing so. So data science, complex data science boiled down to something that people can keep into their heads and use day to day. So that's the really important first lesson. Next is the recognition that in creative businesses, perhaps more than other businesses, but I would argue this is the case in any business. It's people that make decisions, not algorithms or spreadsheets. It's people. I've learned this lesson many times. One time I learned this lesson in the music industry. We had a new artist, a band, we'd just signed. Everyone was very excited. They were going to be the next big thing. No one had heard them yet. I was asked to help choose the first single. There were five candidates. We weren't sure. A great opportunity for data science to make a difference. So we researched the five songs amongst thousands of people in countries around the world. Rich data set, great data science. And I came back with this big smile on my face. I was so pleased that we could help solve this problem because there was one clear winner. One song outperformed all the others as a candidate for the first single. And the guy looked at me and said, well, that's not the first single. Then I said, hang on. He explained that the band weren't ready for a hit. They had some work to do, but also they wouldn't sustain the first hit if they went out with it. They needed smaller songs to grow a niche audience and develop themselves and their audience so that when the hit came along it could grow from it instead of becoming a one hit wonder. Now every data scientist I meet approaches problems in the way I did, which is to find the optimal solution to make the recommendation to tell people what the answer is. It's how we think, but it's not how it needs to work. In a creative business, you have to add skills, judgment, expertise, gut instinct alongside data to make the best decisions. We would have ruined that poor band's career, but combining data with expertise made the best decision and gave them the best chance at success. Now, one challenge we have as data people trying to help creative businesses is, is that people aren't hungry to be helped. They're not standing up here saying, please come help me, I don't know what I'm doing, I need your data. But we're in a fortunate place because these days those exact same people are using data to help them make all sorts of other decisions. There is a precedent, whether it's do I need an umbrella today? Where should I invest my bonus? Or do I need some steps? Or can I send my assistant to Starbucks? For me, they're using data to help them make decisions. And this is the key for how I found it works in creative businesses on the business side, which is give people data visualized brilliantly that helps them to make a decision that doesn't tell them what decision to make. None of these tell you the answer to your question. They help you make great decisions really, really easily. And for me, that's been the key data tools in the hands of people. I care about this a lot. As you might be able to tell, it's because not only do artists and authors and TV producers need help, not only can they benefit from data and data science, but also I've seen so many great data scientists, so many great tech startups in this space, get gigs at companies I've worked at and fail because they didn't understand one or more of the lessons from today and they quite often get unceremoniously booted out. It's a tragedy. So I want data science to play a big role in creative industries. It can play a big role with some pretty simple lessons. I've seen that if you take, if you remember that it's about the people and you give them data to help them make decisions, then data can be at the heart of the creative process and it can help people to make better, bolder and faster decisions. Thanks.
