Disco: A place where data comes to dance

After a modest thumbs up from my dad on my last post on how we set the foundations for people analytics at Spotify, I’m now encouraged to share more on how we’ve created a consumer experience in connecting the business to data content that we know they want to view. 

It’s a tad cliché to start a blog post with a quote, but in the spirit of embracing the cliché, Steve Jobs put it best when he said, “Let’s go invent tomorrow instead of worrying about what happened yesterday”. That ethos has been at the core of people analytics at Spotify. We aren’t defined by industry standards or best practices – we are driven by redefining the space. Except, of course, with regards to data protection and privacy – it’s 100% industry standards and best practices throughout that side of our data work! What this bold claim is really about is the packaging up of already available data in a way that a HR/People Analytics team wouldn’t normally consider. It’s the core principle of having the right mindset, and using our imagination to open the door for ourselves (and hopefully also for others) to explore what’s possible.

Radical transparency and the consumerization of data

At Spotify we believe in radical transparency, and not just internally. It’s a humbleness grounded in our Swedish roots. My interpretation of radical transparency is how we share our experience in order to raise the bar for everyone, so that we can continue to drive innovation together. 

In the interest of that transparency, let’s say it up front: my background is not HR. When I talk about driving innovation, I take more of a generalist approach and think about analytics in general, not just people analytics. This is relevant because the approach we’re taking here in Spotify people analytics is one of the most important developments not only in people analytics, but any analytics function, in any business. That development is the consumerization of data.

Some would argue that consumerization of data has already happened. If you look at the plethora of business intelligence tools flooding the market – with their endless features and slick-looking user interfaces – there’s been a significant shift in the offering. The reality, however, is that they largely service those who’re already strongly data literate, not your average business user. It’s the thin end of the wedge.

Take a moment to consider a truly consumer-led approach to consuming data with no prior knowledge required. Imagine a reality where you weren’t constrained by the quality of your data, or the skill set of your analysts to produce a visually great dashboard, or your own ability to interpret data. We hold a genuine belief that we are stepping into a new dawn for businesses everywhere. “How?”, you might ask.

Some advice I was given years ago was “beer for creativity and coffee for execution”. The next stage was the product of one too many beers and some pondering thoughts, and we think it’s going to change the landscape of how we interact with data. The pondering was over quotes from our CEO, Daniel Ek. Within a multitude of quotes there is this one “Spotify is a platform; it could be expanded to other types of content” and that got the cogs turning! 

If we oversimplify what the Spotify consumer product does, it essentially connects listeners with audio content that they love. The secret sauce is how. We’ve all had that experience of opening the Spotify app and either seeing new music from a favourite artist, or a suggested song you’ve never heard of, which after that first listen then gets played on repeat for the next month. We started wondering – what if you could replicate that experience but with data? 

That was the challenge we set ourselves. We set about creating a truly consumer experience in connecting our customers (by customers, we mean everyone across the Spotify organisation) to data content we know they want to view, and additionally to suggest content they didn’t even know they’d be interested in. Effectively a Spotify for Spotify. We call it Disco. It’s a place where data comes to dance.

Time to Dance

The focus of Disco is twofold: (i) a data warehouse capability that rebuilds the foundations to establish a platform that allows us to progress to the next level, and (ii) a user-facing app that reimagines how we consume data products.

As a continuation of The Water Principles, the data warehouse capability has been an exercise in going back to basics. The simplest description of the strategy behind this is if we imagined a single line of data for every Spotifier. From the moment you apply for a role to the moment you leave. It would require the aggregation of numerous disparate data sources that were ultimately held together at the individual employee level.

What’s important to mention here is that in bringing together these data sets, we’re not talking about the creation of some weird big brother state that wouldn’t be out of place in a George Orwell novel. The strictest standards in the handling and privacy of employee data is our number one priority, as it should be for any people analytics function. Employees are never identified at an individual level in the outputs of Disco and nobody has access to data or information that they are not authorised to see. Whilst I mentioned the premise of the consumerization of data, that does not mean access to all, it’s about how people interact and navigate the outputs.

The second part of Disco is the user-facing app to consume and interact with the data. Much like the warehousing capability, we’ve approached this with a completely different mentality than traditionally adopted in this space.

 The specific challenge that we addressed is that stakeholders across Spotify did not have access to the people data that they needed/wanted because there was no consistency in output. The people analytics team was bogged down in transactional requests. Back to my earlier point about thinking about the industry more broadly, one of the biggest challenges facing any data analytics team is that they seldom move past servicing transactional requests or solving data challenges. What if we could build a platform that finally solved this and fundamentally changes the way this space works?

People analytics as a strategic partner

What Spotify does fantastically well is connect listeners with content that they love and can be inspired by. The challenge with data outputs is no different. The heart of the Disco app is connecting users with data content in a way that is intuitive, personalised and ultimately gives them access to the information that helps them to be more effective.

In taking the conceptual idea of a Spotify for Spotify, we take the core premise of a song, album and playlist, and flip it to rebuild the foundations to metric, data source and dashboard. The experience of Disco is much like our core consumer product and the bet is that the familiarity will help address the challenges we currently experience with regards to a user’s ability to find the data outputs they need.

More fundamental is the aspiration to bring about a new way of thinking with regards to the consumption of data. The often-difficult skill set to find in the analytics profession is being able to cleanse data and produce great visualisations. This is commoditised in Disco, removing our reliance on this skill set, whilst also providing a consistent experience for the user. The traditional need for building dashboards (another under appreciated skill set) is replaced by a focus on curation. In the same way we would curate an audio playlist, the focus is now on finding complimentary metrics that tell a story on a particular topic. We’re moving away from the transactional to being a strategic partner. A partner that can work with stakeholders across the business to expertly provide a new lens on solving our biggest people challenges.

I appreciate that although this is interesting, it’s difficult to imagine the reality. In the interest of radical transparency (but with integrity and respect for data privacy) we’re working on how we can demo some of the specifics but in the meantime, we hope this provides some inspiration for daring to be different, raising the bar of what is possible and allowing your people analytics function to really become a strategic partner. 

To end as we started: “Think left and think right and think low and think high. Oh, the things you can think up if only you try.” – Dr. Seuss