When I was a kid, my Grandpa bought me a hardback tome called the Guinness Book of Everything, a book of facts and statistics about everything from the world’s busiest airports to the longest official country names. I loved it.
Consuming bitesize pieces of ‘knowledge’ from such a breadth of topics was addictive. I didn’t know which world I’d be thrown into from one day to the next, and I relished the prospect that there was so much more to discover in the pages to come.
A false sense of discovery
Today, the idea that a book of facts would be a page-turner seems almost ridiculous. We have an unprecedented ability to discover new things, on demand, at our fingertips.
However, the awakening brought about by the Internet is being tempered by its very own lifeblood: data. The proliferation and use of personal data - preferences, likes, views and so on – to personalise content and experiences has created a new, illusionary concept of ‘discovery’: the “you might like” look-a-like recommendation, as pioneered by Amazon and carried forth by current flag bearers like Netflix.
Granted, such suggestions are rooted in millions of hours of viewing data from people just like me. They’re not wrong, but in some ways, they’re almost too right. They send us down a specific, familiar path, keeping us blinkered from the weird, the different, the things you never knew you’d find interesting. They’re selling us short on the beauty of the world-wide web. Why doesn’t Netflix surface the weird patterns in the data, not just the obvious ones, throwing us a “you might not like…”? By being exposed to new and different genres or topics we could become stickier, more profitable customers.
Avoiding a blinkered approach
There’s a case to be made that we can be guilty of the same blinkered approach in our industry.
We’re awash with data, and feel compelled to use it, with varying levels of success. Yet with any sort of data-driven insight, we have an incomplete view. We only see the small parts of a person that interacts with our machines: a song we play or a photo we like. This patchwork picture is often worse than no picture at all. It reverts us to use stereotypes and broad segments to simplify the complexity and encourages us to adopt the “you might like” approach.
Telling someone what we think they want to hear, or showing what we think they want to see, is a finite strategy. Eventually we either run out of relevant content or, more likely, our customer runs out of energy for the type of content we’ve been furnishing them with.
Our industry is about inspiring new customers and constantly reinvigorating relationships with customers we already have. Success comes from continually spiking people’s curiosity and that requires adding a bit of idiosyncrasy into our planning. This isn’t about being random for the sake of it. Clearly that is not an efficient or defendable strategy.
But looking for adjacencies or anomalies in the data, borrowing models from other industries, or even simply questioning exactly what business you’re in, can unlock more than just our customer’s curiosity, it can uncover a whole new line of business. Dollar Shave Club applied a subscription model to a product traditionally treated as an FMCG by the industry and created a billion-dollar company.
Celebrate the unexpected
The internet was originally touted to be the home of the curious, and now it seems to be at the forefront of limiting our curiosity. Relying on algorithms and look-a-like models alone limits the opportunities for brands to inspire and engage. That doesn’t mean we should ignore the data we have in favour of a totally serendipitous approach, but we would do well to remember that discovery can just as often be driven by the anomalies as the trend.
A “you might not like” recommendation would not be a bad place to start.
Andy Stern is the strategy director at Isobar.