Data Storytelling, Part IV: Know Your Audience
My last post talked about knowing what not to say about your subject. In that post, I spoke to my audience: you. We imagined that you also had an audience – a layman friend. I said, “don’t distract her with all of the stuff I ran through in this blog post. It doesn’t matter to her.” The information in that post was appropriate for you, but she’s a different audience, and she didn’t need as much information as you got.
Implicit in this is a very important point that we’ll explore in this post: Before you really know what your story should say, you need to know your audience.
We’ll start with a familiar audience and then discuss a less common one.
Easy target: the CEO
Your first job as a data storyteller is to listen. Pick out the nouns your CEO uses when she speaks about the business. These nouns are your potential characters.
When a CEO looks at her company, she sees an array of organizational “characters”: finance, product, services, HR, support, sales, and so on. These characters are contextualized by the market, which adds additional “characters” to your story, such as competitors, customers, partners, and influencers.
These characters and their behavior (which will be events in your stories) are what make the company interesting to the CEO. You know that because they’re the characters she tells stories about. That means you should tell stories about these characters, too, and not others.
Obvious? Maybe. But it can be harder than it sounds.
Say you work for a shoe retailer, and you have a chart from the retail sales department that indicates a small but noticeable cost-per-sale increase in stores. It’s the kind of thing that might show up on a dashboard that conveys ordinary status, and we know that a status isn’t a story.
Attempt 1. Suppose that, in the spirit of storytelling, you show that chart to your CEO alongside other details from the retail sales department’s dashboard. Those details include people, departments, and entities that she normally doesn’t talk about: Store 896 is an outlier, having staffed up recently; Store 891 is paying a lot in overtime.
You’ve beefed up the data being conveyed, but you’ve missed the mark in two ways.
First, you’re not telling the CEO a story that involves her characters. She could reasonably ask why you’re showing her the VP of Retail Sales’ dashboard.
Second, you’re not presenting her with finished analysis. You’re presenting the data you needed to do your job, but you haven’t done the analysis needed to give her the picture that she needs – a story starring her characters.
Attempt 2. Let’s try again, still supposing you start with that chart showing a cost-per-sale increase in stores: After all, you want to get her attention by starting in medias res (“in the midst of things,” as discussed in a previous blog). Now supplement that with information showing cause and effect, personal impact, and insights she currently won’t get from anyone’s dashboard.
Start with a “Star Wars scroll” of context, populated by characters that your CEO knows well: market data (sourced from marketing) indicates a customer preference for trying on shoes before buying, staffing costs (from HR) are relatively steady. Online sales are up (from sales), but there’s a big uptick in buy-online, in-store pickup sales.
What happened? Real estate costs (from operations) are starting to increase as the company starts to renew leases that were originally taken out in that big expansion ten years ago.
What could happen? There are different possibilities. Discuss the effects of (say) going online-only, cutting real estate costs by half to support mainly in-store pickup purchases, and sticking with the status quo.
Now, through your story, she understands what was, what is, and what could be, and she can make more-informed decisions.
I’m not a shoe guy. I don’t want to guess what the right decision would be. But if you know the context, the characters, and the interplay of cause and effect in your business, you can do the analysis and tell the story that helps the CEO make her decisions.
Not so easy: the truck driver
In my experience, people don’t think about truck drivers much in a BI and analytics context. As a result, this section may be harder for you to dig into.
If you don’t have truck drivers, think of a Truck Driver Equivalent. A TDE is a role that’s relatively common in your company, but one that will never be expected to use a BI or analytics tool in the normal course of business. Every company has a lot of them.
If you think of data and analytics in terms of “users,” this might not be a comfortable moment. Users of data science, analytics, business intelligence (BI) – you name it – are supposed to use the data science, analytics, or BI tools, aren’t they? That’s why they’re “users.”
I have to disagree. That’s the mentality that keeps adoption rates of BI, analytics, and data science (N.B., not tools) below 25% in so many organizations. Truck drivers and other TDEs make up the 75%, and as an industry, we tend to ignore them.
So clear your mind and think of truck drivers.
As before, your first job is to listen.
The characters in a truck driver’s story might include the truck, other truck drivers, warehouses, loading docks, and customers. Events in their story might include traffic incidents, on-time deliveries, damaged shipments, traffic tickets, and refueling stops.
What stories can you tell truck drivers, using these characters and events? Which of them are valuable?
Remember that I’m not talking about dashboards. Truck drivers are used to dashboards, and could use them if they’re well-built. They could benchmark their delivery performance, attendance, and traffic violations compared to the company average, for instance. They should be “users” in that sense already.
(Are they? When was the last time you focused on giving basic information to your truck drivers and TDEs? When was the last time you talked to them to understand what information they need to maximize their performance? When you talk about “user adoption,” do you include them?)
For storytelling, you want to use data to open them up to new information and change their behavior. Here’s one possibility: Show your truck drivers the adoption of new picking equipment in certain warehouses, the decreased average turnaround time spent on the loading docks there, and the increased per-mile payments they’re set up to make each day.
That’s new understanding, not just a dashboard, and it can affect their choices: Their willingness to adopt new procedures to accommodate the new equipment may be a lot stronger, and mistakes less common.
A freshly opened can of worms
This post was all about knowing your audience. The next one will be about knowing your characters. These nouns that we’re gleaning from business sentences so we can identify characters? They’re master data – which means master data management is an important part of any discussion of data storytelling.
See you next month.