I just want to thank Anna and Jeremy and Formulated.by team for having me. I’m a designer as it says on the slide so I’m an ambassador from the other side. We’ve been hearing a lot about how data is about communication and how it’s about teams. I am here representing a team that’s under-considered.
You probably don’t think much about design teams. Often, we’re like the Brooklyn Nets of your organization, down in the ranks. In fact, I was telling my mom that I was coming here to a data science conference and she asked, “Do you think they want to hear from designers?” I told her that I didn’t know and that we’ll find out. So here I am and I’m really excited to be here with you.
Specifically, I am a product designer. I work at a company called The Black Tux in Santa Monica. Not only am I a product designer – I am a designer who focuses on growth and if any of you know anything about growth, you know that it is founded on data. By the nature of my work, I have gotten incredibly involved in data and I’ve worked with all kinds of organizations. I used to be a consultant, so I’ve worked with large companies like Toyota and Macy’s and then I’ve also worked with startups. I’ve noticed gaps at all of these kinds of organizations and that’s what I’m here to talk about today.
I’m seeing this disconnect happening between data teams and everyone who’s not on data teams. A while back, I worked at a startup in this scenario. As a product person, and I went to our data scientist and said that I’d really like to know why people are clicking on certain ads but not others. He said that he didn’t have time to explain because he was making revenue charts for the CEO using machine learning. We were at a standstill because I couldn’t get the information that I needed to do my work. Moreover, I didn’t care about his projects because I couldn’t make it actionable for my work.
As I mentioned before, I’ve seen this over and over again in organizations and as we’ve heard today and yesterday, I’m not the only one that’s experiencing this. I’m seeing that there’s not a shared language between technical folks and non-technical folks. If the non-technical people can’t get the information to do their best work, they can’t improve the business and if the technical people right can’t connect with their work, then nobody within the organization will value the data. And then, if I’m in leadership at that company then I don’t hire more data people and I don’t leverage the data I have. I’ve seen data teams just completely go away in those scenarios and so I’m here to plea to you to help me fix this gap.
One thing that I see when this disconnect happens is that people develop their own solutions. In other words, the marketing team grabs Google Analytics, everyone’s favorite tool. It’s actually my least favorite tool. The support team has their own built-in analytics because no one let customer support in on the data so they just went ahead and built it. Then, the data team has their BI tools and they make their own tools so I really would like to encourage you to think about how people on your team experience data. I’m approaching this through the lens of experience and I’m encouraging you to do the same.
I have three sort of practical buckets of ways that I think you all can help bridge this gap and as the people with the data, you hold the keys. I’ve been giving this talk to designers and product managers who want to get in on the data, but for those of you who are on the technical side, there are many things we’d like to see. We need to be having these conversations about who is in charge of reporting things and who is in charge of teaching others how the system works. It’s critical to be having these conversations and then to make that really clear when people join your team.
See talks like this in person at our next Data Science Salon: Applying AI and Machine Learning to Retail and E-commerce, in Seattle.
Some of the ways that you can develop these roles are through one-on-ones. You can just hold one-on-ones with different people who you think might be interested in helping. You can also make better visualizations with more rich information about what people actually do what roles they serve. As a person coming from startups, I’m very sensitive to having way too much to do. If you try one thing to make roles clearer at your company, you will be ahead of the game.
The other thing that we can look at is passive information. As data practitioners, there’s a lot of information that is communicated through interfaces and through language and this can be good if you prefer not to talk to people. I see a lot of situations where people don’t trust the data. They can’t follow the trail and they don’t know who to ask. Or maybe, they can’t see in the tool the pathway to get to that raw data. First of all, be sure to choose tools that the team understands. You guys have all seen some gnarly data and it’s really not cute, so think about the tools that you’re choosing for people to look at reports. Think about the communication tools that you’re using and the platforms that you’re using. We want to intentionally show all that data and make it dummy proof.
When humanizing your data and making it really clear, just remember that you’re dealing with people. My expertise is on the behavioral analytics side and I’m constantly blown away by what people name their behavior. Especially, when people don’t understand the data, they just make their own thing and send it into the system. So in making things, I want to make things dead simple and be sure that this is what I would say out loud to someone. I’m going to call my data elements what I would call them to customers looking to buy them. That way, we have a common language across the organization. All of the details of the data need to be transparent.
When you have your tools and you have your humanized naming conventions, you are well on your way to this beautiful system that everyone in your company can read and understand. Another thing that we could all do and specifically any of you who work on the data platforms can do is make it easier to put the definitions and instructions like right there next to the deliverables. For example, Amplitude, my favorite behavioral analytics software, has taxonomy built into their platform. If I hover over something, it tells me what it means. It’s a recent addition though, and it’s certainly not built into every software. If you guys have proprietary tools, you have to build this in so that people can get involved in the data right away and actually use it in their day-to-day work.
Lastly, we need to think about how can we build such capabilities within our organizations. You don’t have to be Facebook to do that. I’ve worked at ten-person companies that have been doing this. So how can you help level up your teammates? Most of the time, it’s just that they’ve been presented with some sort of confusing lens on the data. They’ve been told that it’s not their problem, but also know it doesn’t have to be this way if everyone valued data in their work. The fact that not everyone does is soul-crushing and it’s bad for business. We all know that data is valuable for businesses and that’s why we’re all here so we want our teams to understand and use data.
Airbnb has their own internal data university. I don’t know if anyone here has been through that, but it sounds super cool from the outside. They will train you at different levels to use your data so everything from literacy to the actual programming they’ll teach you. I think some people genuinely become data scientists through this and it’s a great model to look at.
As I said earlier, if you’re at a smaller company than Airbnb, start with a lunch talk or a slack message or even a blog post. Anything you do is better than nothing so just think about how you can include people and how can you reach out to the marketing team. You can always check-in with your team members to see what data they’re using and how much they understand about what you do. The whole idea is that everyone is empowered to help each other, so it’s not just about your job. It’s also about the value you provide to the greater organization, which happens through helping others.
How can we help each other strengthen the value of data within our organizations? Encourage people’s questions! Let people ask dumb questions that don’t make sense to you and empower your teammates to answer these questions and to be available for this kind of stuff. These little entry ways allow people to get involved. When people know who to talk to, they are way more empowered to do their work.
I just want again to implore you to think about how your team experiences data and help them get more involved. It’s up to us to fill this gap. I’m trying and I’m here to encourage you to try as well. I would love to talk to anyone about how to do this or specific challenges you’re facing. We all want to value your work and moreover, we want you to be working on complex projects rather than the basics. Let’s get this foundation there so that we can all do more interesting work. Thank you.
For more content like this, don’t miss the next Data Science Salon in Seattle, on October 17, 2019.
