Guide to telling stories with data: How to share analytics insights
A comprehensive collection of articles, videos and more, hand-picked by our editors
You're investing a lot of money in business intelligence tools and applications to keep up with your organization's changing business needs -- perhaps hundreds of thousands of dollars, if not more. But are you investing enough in the people who you expect to use those tools -- or in the right kind of people to begin with? Probably not.
In many cases, you're still primarily looking to fill BI developer or dashboard developer positions. As BI leaders, we need to put an end to this way of thinking about business intelligence skills, which is as outdated as the tools where all of this "BI developing" has gone on for the last 20 years. As data becomes more valuable, so do people who know how to use it effectively to help drive business decision making -- and they're not developers in the traditional sense of coding reports.
If you're in a more forward-looking organization, you're trying to move in a different direction. You push members of your team to use data visualization best practices and tell stories with data to help make the results of BI applications more understandable to business executives. Most people, though, don't have the required skills -- or they have some but not the full combination that's needed to effectively present information to execs, let alone become future BI leaders themselves.
Modern BI skills in short supply
Whether you fit into the first or second scenario, you need to start (re)building your team to compete in a world where data storytelling and visualization are in the spotlight and the underlying data now has more of a supporting role in decision making, often behind the scenes. In five years, you'll be able to do so by hiring skilled data storytellers from outside your company. Now your only real option is to "create" the workers you need in a laboratory (of sorts). You can do so by following these three steps:
- Identify people with the potential to learn and execute the new BI techniques.
- Train them to gain credibility in the company, and continue to build on that foundation.
- Invest what's needed to make them leaders -- if data is strategic, so are these folks.
This article, the first in a series of three, addresses step No. 1. The challenge is that you can't easily predict who would make a good candidate. There isn't a checklist of technical skills to mark off; it's not necessarily your "highest performer," whatever that may mean. What's most important in the kind of person you seek is a set of softer skills: comfort with ambiguity, a sharp attention to detail, the ability to ask good questions and listen well to responses, and confidence in being creative.
It's likely you're scratching your head and wondering what those four skills have to do with business intelligence. In fact, they're all predictors for being able to understand the business -- or, as I like to say, being intelligent about the business. Learning them is much harder than learning how to use software, conducting an analysis or creating a data visualization. But they're the keys for the next BI superstars in your organization. I'm asking you to make a mind-set shift.
Game on for finding the new BI breed
So, what can you do to help identify the right people? My proposal is going to make you smile or cringe -- or maybe both. Hopefully, though, you'll find it practical, and fun. You're going to set up a competition around a BI project in which results are measured and the marketplace of customers -- for example, internal business managers -- decides the winners. This is how it works:
- Select a project. You want the work done as part of the competition to be immediately usable for two reasons. One, you'll be able to easily form a customer panel if business execs see how the project will benefit them. And two, you want the project's success to be visible and serve as an example of the kind of impact that's possible with a new approach to BI oriented toward visualizing information and telling compelling data stories.
Pick a project that isn't mission-critical but is meaningful to your organization and generates data that will be used frequently. The project should be done in stages; that way, you can adjust the next deliverable based on feedback from the panel. And specify the time frame for each stage so the competitors have a clear picture of the schedule and can make a solid commitment around their ongoing work. They should all set aside the time to participate equally; for example, you could have them book their calendars to protect project work from other priorities.
- Form a customer panel. Have a small group of execs -- three is a good number -- sponsor the project and select the winners. Once the panelists have been chosen, meet with them and agree on the end goals of the project. What questions must be answered; what business decisions can be made based on those answers; what actions should be enabled?
At each stage of the project, individual competitors should be given about 15 minutes to talk to the sponsors and get input on their work. That also gives the sponsors time to interact with them. Provide the panel members with a general idea of what you want to hear about the interactions, plus a cheat sheet with prompts for taking notes on the four predictors I outlined above.
- Call for (all) competitors. More participation is better. You'll build excitement, generate more ideas and possibly find a larger pool of potential new BI team members. Open the competition to as many candidates as reasonable, and be careful not to limit your thinking on who is a good one. Prior BI or analytics experience isn't necessarily required.
Because of that, you should provide the competitors with access to an existing BI resource who can handle the software used to prepare data and build visualizations. That way, not knowing how to use the software won't be an obstacle to making their ideas come to life.
- Measure the success. The sponsors must be unanimous in their decisions. That's why three is a good number to encourage some discussion, but not too many to make the process unworkable. If multiple competitors are deserving, it wouldn't be a bad thing to have more than one winner, and there should be some silver and bronze awards in addition to the gold(s).
Of course, beyond handing out awards and contributing to a business initiative, you're looking to assess the skills of the competitors with an eye toward future BI roles. In addition to weighing them against the four predictors, you should grade them on how correctly they present the data. Getting it right is critical to building trust when you're relying more on visual patterns and less on rows and columns of numbers.
With traditional business intelligence skills becoming less valuable, my proposal encompasses common assessments you'd have to conduct to identify workers who have the foundation for flourishing in the new world of BI -- albeit with a twist. Even if you aren't able to apply these ideas, you can benefit from rethinking your current BI approach and looking for opportunities to deal with a new problem in new ways. Now go find some potential superstars and start nurturing them. Excelsior!
About the author
Lee Feinberg is the founder of DecisionViz, a management consultancy that helps companies escape the legacy of reporting data by transforming complex data into simple pictures for making decisions. He also is a frequent presenter at domestic and international events. Email him at Lee@DecisionViz.com.
Which BI skills would be best to have?
Why the right analytics skills can compensate for tech know-how
Data storytelling advice from an expert
BI development can make your business run more efficiently