Data power to the people
How to make your non-tech colleagues use data interactively.
Many companies aim to be ‘data-driven’. Whatever your preferred definition of that term is, there is a lot to read online on what you should or shouldn’t be doing to get there: Smart KPI Setting! Model-based decision making! Be data-informed, not data driven!
This blogpost won’t help you define what data-driven means for you or your company. It won’t tell you what metrics to track, what data-sources to scrape, or what dashboard to put up on a big screen in your office. While all of the former is important, at ParkBee we think that if you are only tracking KPI’s or looking at dashboards, you’re missing out.
Part of our job as a data team is ensuring that everyone can get the insights they need to do their jobs. That’s why this post will share some of the lessons we learned at ParkBee in making sure our team members can not just look at data, but interact with it, too.
1. Empower your team members to DIY
A good starting point for using data throughout your company is to make it visible and available using BI-Software. There are a lot of tools out there and you’re probably already using one. Without going into detail and comparing them all, we think that a good tool should at least allow for the following:
Everyone who looks at a dashboard for their job, should also be able to build, edit and share dashboards themselves, too.
Building a graph or dashboard from scratch shouldn’t just be something a Business Analyst or Team Lead does. Instead, we allow and encourage everyone to copy, edit, reshuffle and share each-others dashboards. By adapting existing graphs and dashboards, users get a sense for what is possible with the data. What’s more, they can see how to structure the data to get them the answers they need. Ultimately, we want to train everyone at ParkBee to think about business questions in terms of queries: “what set of dimensions and measures do I need to get the insight I want?”
We do this because in working with data, we think it is important to keep asking “Why?”. For example, after taking a look at a KPI dashboard, you’ll find yourself asking “why are we performing like we’re performing?”.
Allowing users to answer these questions themselves is vital in a good BI setup. We think this is not just a time-saver for your data team. We think that allowing the people with the domain knowledge to do it themselves, ultimately leads to better results.
2. Build tools, not slideshows
In the end, the right BI-software and setup will never be able to handle everything. There will always be custom work needed to ensure that you get the most out of your data. This is where your data team comes in: helping other teams with the more complex questions and challenges. Over time however, we noticed we were spending more and more time on small tasks. Think of things like adapting previous analyses to new cases, re-running a python script to get updated data, tweaking that one graph, etc.
To avoid frustrating ourselves with tedious and boring work, and our colleagues with waiting for simple stuff to get done, we ended up implementing the following basic rule:
The end result of a request to the data team should always be reusable.
Imagine a colleague needs help doing a pricing analysis for a certain location. Instead of providing them with a pricing analysis in Google Slides, for example, you can also host an interactive jupyter notebook, allowing your colleague to re-use the analysis for other locations as well.
With the right basics in place, making analyses more widely applicable and available takes only a small amount of extra time. Things like creating interactive notebooks, creating a small flask-web-app, hosting an interactive folium map or even helping users to build their own spreadsheet using tools already available to them can save you a lot of time in the long run.
3. Be approachable…
Encouraging your colleagues to interact with data themselves also means you’re asking for their time and effort. At ParkBee, we make sure that everyone we hire has the right mindset to use data in their day-to-day. However, with all these tools available, it can be hard to know how to translate your ideas into actual analyses. That’s why we encourage our team members to come to the data team if they’re stuck or unsure about their results.
These small interactions help us to stay up to date about the data needs of business, and give us the chance of avoiding requests for things that already exist. It also allows us to help ensure that our colleagues draw the right conclusions from the data.
4. …but not too approachable!
In the end, building and maintaining internal data tools is not the only job of the data team here at ParkBee. We quickly experienced that it is important to strike a good balance between being approachable and having enough time for our other tasks.
Generally, we encourage our co-workers to get help from their own team-members first. This usually filters out the smaller questions like “how can I adapt this graph in our BI-software” or “Where can I find our KPI dashboard?”.
If they can’t figure it out or no one is available to help, they can just come straight to the data team: no need to plan a meeting or wait for open-office hours. Usually, we can help them out right away with the tools we already have in place. If not, we ask them to put in a ‘formal’ request, clearly outlining what they need. While this can be a bit of a hassle, it helps to determine if a feature is really vital or just ‘nice to have’. If it is the latter, usually it is too much effort to write out the requests, and we won’t hear back.
It took us almost a year to put all of the above into practice. Now, everyone across the company can use data to help them make well-informed decisions. This has changed the way we set expectations and evaluate results.
Of course, the lessons we’ve learned aren’t the be-all and end-all of good BI. We’re still learning more and trying to improve how we work with data in our day-to-day. That also means knowing where your analyses fall short, where you may be biased or where your data is lacking. After all, if you want to be data-driven in your business, you’re going to need the right people to build and use these interactive data tools.
Sounds like you? We’re hiring!
I work for ParkBee. We develop smart tech. Our Mobility Management Solution optimizes the use of underutilized parking space to get cars off the street and KEEP YOUR CITY MOVING.
Check out ParkBee’s Careers Site , or follow our @lifeatparkbee Instagram account.
