Q&A: The Data Science and Design Behind Our New Look
Chartbeat’s Design Director and Product/Data Chief discuss the work behind our new user experience. Interviewed by Nalini Edwin and Jill Nicholson.
To start — Renee, could you walk us through the context behind this work? Through some of the design thinking behind it?
Renee Solorzano [Design Director]: A constant project of mine is to keep moving us in the right direction with UX consistency and a cohesive product. Certain parts of our product were independently built to address specific problems, which was good in that we were solving problems that needed to be solved, but overall led to a cross-product user experience that really needed more consistency and smoother user flows. We were building a lot, but needed to step back and look at the bigger picture.
So our goal on the product side was to begin weaving together an experience that brought all products to the same playing field. We wanted this push to ground us so that, moving forward, we are constantly thinking about the platform as a unified whole. In terms of the interface, we were exploring how to evolve from what we had, making it look cleaner, bolder. But the evolution didn’t really have a story. So that’s really what the rebrand solidified — the story behind how we present ourselves.
What do you feel is core to Chartbeat’s identity, and how do you feel that it comes through in the enhanced product?
Renee: For me, it’s being clear and intuitive — and in order to do that, you have to understand your user. I love that we are a partner with our customers — we’re always visiting and talking to people to understand their workflows and building our features and products to suit them.
That was also apparent inside Chartbeat when we were interviewing our colleagues. One of our PMs reminded me that we’re “the only analytics company in the publishing softball league” … things like that. That’s always differentiated us. It’s reflected in the product in the copy that we use, the user experience, and in making everything as clear and concise and edited to really get to what is important to our users. It’s something I feel we’ve always done well, but this rebrand makes it even crisper.
Josh Schwartz [Chief of Product, Engineering, and Data Science]: From a product perspective, just to expand on that, I think Renee touched on how our focus on media means that we can deeply understand what’s happening in the industry at a high level, but also from a workflow perspective, so we can truly focus on building what actually works for people.
The other side of that is that analytics products are only as useful as their usage. Right? If you make a product that people don’t love to use, it doesn’t do anything at all. And so we’ve always been focused on understanding what people actually want to do with analytics, and then making a product that fits those needs. I think a lot of the work in the rebrand is about making adjustments that make the product work even better in sight of people’s workflows.
We’ve always been focused on understanding what people actually want to do with analytics.
What makes you think differently about how we guide people through the product?
Renee: Data. One of the biggest things that we saw in the data was that not many people were navigating to other things besides the Real-Time Dashboard. We want, going forward, for our product to be thought of more as a system, rather than: This is the one thing that you know us for, and this is the one thing that you will get from us. It’s more like: How do we put our efforts into making sure that our data is important beyond real-time, that we can surface the benefits of context in the right places, and that the depth of Chartbeat can be used across the organization?
Josh, how do you see data science fitting into that system as Chartbeat begins those efforts?
Josh: Data science has always been core to decisions about what we build. One thing that’s unique about Chartbeat is that we try to deeply understand any piece of data before we put it in the product. The first phase with any new thing we’re going to incorporate into the product is a data discovery of: How might this look? When we built a headline-testing product, the first thing that happened before we wrote a line of code was an analysis of how different headlines perform — the approach was to understand that, then run simulations, and so on.
That desire to deeply understand stuff, rather than just report things without really understanding them, is essential to what we do. Just like we would never make a product without understanding how people use it, we would never make a product without really understanding the data that went into it.
Data science has always been core to decisions about what we build.
Doesn’t that go deeper, too? You guys are thoughtful not only about what goes in our products, but also about what doesn’t. How do you make sure that we don’t data overload, and that we’re surfacing the right things? What are some of the discussions that you have when those moments come up?
Renee: There are two things. One is asking: What is the data that’s most important to people? So when we’re doing design explorations, we’re actually showing lots of different options to people and seeing what matters to them most. We have a lot of different customers with a lot of different use cases, so we have to figure out what are the most important pieces across the map that apply to the most people, but also: What is our point of view on what is most important? So it’s a balance of both.
And then from a user experience perspective, it’s: How do we make sure that what we’re putting on the page isn’t overwhelming? So we think a lot about the hierarchy of information, and a lot of decision-making goes into what we actually surface on the page.
Josh, one of the big challenges for the data science team is that there are so many different sources of data to potentially combine — social data, app data, all those things. For you, what has been something that’s interesting (or difficult) in thinking about how to providing a perspective across so many different content distribution channels?
Josh: I think there are two challenges to incorporating data from the wide world of everything, from social sharing to app traffic to offsite consumption to onsite consumption. One is from a measurement perspective: just making sure that you’re measuring as accurately as you can in as many places, understanding the weird quirks of how you can measure insight of an instant article versus a website, things like that.
The other side is understanding what people are actually going to do with it. When you’re an editor tasked with programming Instant Articles, you have different visual tools you can use than you can on the web, you have different decisions you can make, and so on. So it’s making sure that you understand what people are actually going to do with the data and you’re building the product to suit that.
I think a good example of this is our Offsite Social module, which rather than just saying, “Okay, you can break down stories by concurrents. Now you can also break down stories by likes and shares,” was based on our question: What are people actually doing with the social data?
We learned that people want to discover what the [social-media] landscape is. They’re trying to find influencers, they’re going and talking to them, and they’re addressing their audience. People are trying to do social media management for their own accounts, and so we built something that helps them do those things, rather than just saying, “Okay, we’ve incorporated this data in. You can break it down like you can break down anything else.”
All of Chartbeat is really made with the actual user problem in mind. The media is an ecosystem with different kind of needs and habitats; you tailor what you make to fit the part of the ecosystem they’re part of.
To incorporate data from the wide world of everything: measure as accurately as you can, and understand what people are actually going to do with it.
What is your favorite part of Chartbeat, and why?
Josh: I’ll give two answers, and then I would love Renee to give the third. One answer is, I think, the Big Board — because it’s the original thing Chartbeat released that really spoke to this notion of helping to create a culture of data. Seeing Big Boards up in newsrooms across the world, from Egypt to D.C., is something that’s pretty unique and has become this iconic thing representing a newsroom that wants people to know what’s happening on their site. I think that was our first really newsroom-specific product, and there’s something pretty special about it.
I also like Headline Testing a lot, both because there’s good math behind it and because, in some ways, it crystallizes a lot of what makes Chartbeat Chartbeat. On one hand, it solves a real need: we heard from a lot of people who had testing products but couldn’t really use them, because they didn’t really fit within homepage editors’ workflows. So we saw an opportunity to build headline testing right into our Heads Up Display, where homepage editors are already spending much of their day. That is pretty cool. It also helps people learn how to write better headlines, and an opportunity to help people actually hone the practice of journalism is very cool. Finally, it’s very Chartbeat in that it is really centered around the idea of capturing and understanding user engagement, and not just sort of raw clickthrough. That sensibility is very uniquely Chartbeat.
Renee: I don’t think my answer will be better than that.
Renee: Hmm. I think the Heads Up Display is just so specific in terms of its user base, that it was really built with them in mind. From a design perspective, to create something to work on top of any screen, you just have to deal with so many considerations. We had to make something that was so universal, with no idea what would necessarily be underneath the overlay, and to make it work in that environment is really awesome. I’m also super proud of the iOS app. In terms of what you were saying about decision-making earlier: how do you edit down something that is a dashboard, which people mostly see in fullscreen mode, to something that is 400 pixels wide? That was really a design feat, in terms of hierarchy and in talking to users to make sure that what we showed was what they wanted to see, in terms of what kinds of traffic they wanted to stay on top of.
Final question: could you please describe your role at Chartbeat, using only emoji?