H-E-B's Supply Chain and Operations arm has a massive task ahead of us—we need to replace the 30-year-old mainframe that supports our massive supply chain, without missing a beat.
When we kicked off the effort to untangle the spaghetti system of code, we knew that meant juggling the multiple tools our users rely on while we iteratively dismantled the old system. As we investigated what to build in its place, we realized just how many business decisions have been made under the constraints of our existing technology. To smooth the transition, we wanted to create something that allowed our Partners to see the data around ordering products for our warehouses in a new way. What if we could make a tool that solved our users' problems, something that felt both familiar and groundbreaking, and that could serve as an intuitive way to change everything about how our Partners work without changing everything?
We made an order tracker.
That might not seem like much, but our tracker was the carrot that would lead them to a new system. The tool was incredibly tempting to our users, solving a problem they didn't realize they had while getting them accustomed to the interface we'd prefer they work in. It allows Partners across multiple disciplines to utilize data from complementary systems—transportation, order management systems, warehousing, etc.—to provide a complete picture of each purchase order at every step in its lifecycle.
How'd we get there?
Instead of starting with the current problems our mainframe contained, we started at the lowest level, which is all of the data we had for purchase orders. We created a brand new back-end system that aggregated existing messages currently generated throughout our supply chain. With all of that data at our fingertips, we started from scratch and built an entirely new representation of a purchase order.
From there, we progressed to creating a customizable dashboard so Partners can see a streamlined view. Instead of searching for individual attributes, meaningful data is highlighted with calls to action. You'll get an alert if there's an inventory shortage or late deliveries. Different users can take different actions based on the data, but the display works across departments. Best of all, this representation (and back-end) are flexible, giving us plenty of room to evolve.
For our users, this new technology was an easy sell. Who wouldn't want to trade using various individual systems for one condensed interface? All Partners would have to do is type in the PO number and then they'd get the full story of an order. We built a system they'd love to adopt.
The clean interface was key. This tool is used by many different types of users—our supply chain Partners in forecasting and replenishment, our procurement teams (all types of buyers), our Warehouse Partners, Transportation, and Accounting. We needed something they already understood. Everyone's tracked a package (or a pizza), so they can jump right in and not waste time learning a complicated interface.
The goal is to create internal tools that attack business problems by uncovering underlying needs. With dedicated UX designers who can seek out those issues, research the best way to solve them, and inform what we build, we can unlock new ways of working. A sleek interface makes the transition easy—and once new users are engaged, they want more.
When we started our mainframe project, we broke it into three phases. Phase one was to approach the problem from a technical, risk-mitigation perspective—collect and validate the data needed to create a complete picture of a purchase order. The second phase is about cutting the cords that send our data to downstream systems. Phase three is all about taking over order creation and maintenance from the mainframe—and enabling us to break free of some of the current constraints. When we started that first phase, we knew the tracker was our carrot. Purchase orders were super painful—you had to log into multiple systems, which means you had to have access and knowledge of each one. To streamline that into one sexy little view would change our users' lives dramatically.
This little product created a way for us to get our users involved early in the process, to build their trust, and get them to want more. An engaged userbase provides a constant feedback loop. We can evolve data displays and make the product more valuable with each update because our users are on the journey with us.
We've built multiple opportunities for our users to share their needs with us. There's a place to send comments right in the application, so, as our users are in the tool and issues are top of mind they can say "hey this doesn’t look right" or "this would be helpful." We have a product owner that meets with the various groups on a regular cadence to make sure we’re all communicating our progress and priorities. We built a change log within Slack so our users are all updated when we push changes (which is nearly daily), but it does more than just that. Each reminder keeps our evolution top of mind so our users know we’re iterating on their feedback. Hearing from Partners and improving their work has been the most positive experience I’ve built at H-E-B (and I've been here quite a while).
For us, the hardest part is balancing all of these incremental opportunities with the larger goal of fully moving users over from the mainframe. As we have taken in all of this feedback, we have to keep our focus on the bigger picture. Does this small change slow us down or help us more along the way? When we put together the initial strategy, we mainly focused on technology. How do we ultimately turn the lights off on the mainframe piece? We constantly monitor how we can best utilize our time to achieve small fixes and progress towards our end goal. We know the bigger changes will make a big difference, so we have to be sure we aren't delaying their impact.
I've been with H-E-B for 13 years, eight of which I've spent building software, but this tool is one of my most rewarding projects. It might seem small, but it changes our Partners' lives. We're giving them time back by streamlining their processes and making the day-to-day less painful. If their jobs are easier that contributes to better inventory positions in our warehouses, which leads to better inventory positions in our stores, which ultimately makes our customers happier.
Who knew that something so small could make such a big impact?
Hannah Cole is Engineering Manager. You can connect with her on Linkedin here.
For years, customers using SNAP benefits had to pay in store—but when COVID struck, we wanted customers to be able to use their benefits however they felt safest.
When COVID drove massive traffic to H-E-B’s online experience, we realized our work for the big game had been the perfect scrimmage.