As many in CPG circles already know, Walmart is preparing to make massive changes to how it provides data to suppliers – transitioning out of systems that, despite the nuanced interfaces and unique terminology, have become comfortable and familiar over the years. As our CGO is fond of saying, “How do you spell warehouse? WHSE.” 

There is a day quickly approaching where compact store/item-level reporting is going to expand into billions of rows of data across multiple tables, accessed through reporting interfaces or piped through as raw data using API connections. While these oceans of data are going to provide an unprecedented line of sight into one’s business, the thought of wrangling an ocean is daunting. 

Enter data visualization. 

As one embarks on this new stage of The Retail Journey (shameless plug), data visualization tools will be vital to one’s success within the walls of the world’s largest retailer. Suppliers would do well to get to know software like Tableau, Domo, Power BI, Alteryx, etc. and set off! 

Easier said than done, right? Well, fear not! Below are some of the key components of data visualization creation that makes your DATA (at this magnitude we must capitalize the word) work for you, and some questions for developers to ask as they plan builds around each component. Some of the questions may be common across components, but we devs will need the constant reminder.

‍

‍1. Good Data Vis has a Clear Directive

When the user clicks to open a dashboard, it should quickly be apparent to them what the end goal or use case is. There will, of course, be nuances or best practices that will need to be communicated depending on the software used for the build, but there should be as few hoops to jump through as possible between their first click and their end goal. Though not always possible, a page that all parties at a company (execs, sales, analysts, etc) can jump into and quickly glean a helpful nugget is a winning page. 

‍

‍Developer questions: Who is using this build? What are the takeaways that the user will consistently need? Is this a reporting dashboard or a data driven tool? What are the KPIs that need to be communicated for this dashboard/tool to complete its function? What visualizations do the best job at demonstrating the KPIs? What questions does Walmart ask in Line Reviews that this build could help answer, and how do they prefer to see the data presented?

‍

‍2. Good Data Vis has Dynamic Content

It’s not always possible to fit all of the helpful information succinctly into one page, and no one wants to scroll through slicers all day. A good build will contain visuals that tell a story that the user can click through like they’re turning pages. For example, say I am in charge of investigating the root cause of a YOY sales decline in a particular category. A helpful visual story could be a page with a bar chart that shows the total change in sales dollars YOY by category, a line graph that shows YOY trend, and a table with a set of our KPIs. Initially, all visuals are reflective of the past two years of data. I select the category bar that I am investigating, and the line graph that depicts trend over time snaps to show only the sales for that category. I see that the trend began to tick below last year’s trend in week 13. I click week 13 and realize that one of my KPIs (my average unit retail to be exact) is up 12% YOY. Just like that, I am able to report a preliminary cause to my team lead. Dynamic content allows for quick diagnosis and clear storytelling with data. 

‍

‍Developer questions: How can I combine visualizations on a page to allow users to slice and dice with as few clicks as possible? What aspects of our data have cause and effect relationships that I can incorporate to tell a story and answer potential questions. How can I integrate KPIs to help our team diagnose asks about business performance?

‍

3. Good Data Vis Drives Action

Unfortunately, it is all too easy to create a visual that communicates a number, even something as significant as a KPI, that leaves the user with nothing to do but say “yeah, so?” This is a key component of a dashboard/tool that takes it from good to GREAT. For example, say I am looking at my sales performance YOY. A good tool will tell me, “hey man, your sales are down for this category from where it was last year. Here are a bunch of KPIs for you to look at.” A GREAT tool will tell me, “hey man, your sales are down for this category, and that YOY decline is largely due to the fact that this particular item is down 12% in its in-stock from where it was last year.”  Or, better yet, “hey man, your sales are down for this category, and that YOY decline is largely due to the fact that this particular item is down 12% from where it was last year. Here is a list of stores that you should attack with an SSO to rectify this problem.” See the difference?  

‍

‍Developer questions: What is the action I want to drive from the dashboard/tool? Is it a clear path from first click to reported solution/actionable download? 

‍

4. Good Data Vis Embraces Aesthetic

Who says data can’t be pretty? If you’ve taken the time to make a tool that works for your business, take a few extra minutes to make what you’ve built nice to look at. Whether our data brains like it or not, cleanliness and sharp visuals can encourage confidence in the tool/dashboard. More importantly, part of a good tool/dashboard aesthetic is a clear separation of visuals to avoid overcrowding, and a cohesiveness across pages that helps with the flow from first click to action that we discussed in a couple of the other points.

‍

‍Developer questions: What are some simple things I can do to make this page nice to look at (font consistency, color palette, drop shadows, etc.)? Have I tried to squeeze too much into this page? Does this dashboard/tool bring joy to the beholder? 

‍

Here at High Impact, we are ocean wranglers, and we love to help suppliers thrive at retail. Give us a call! 

‍