Working at GoPuff, an instant-needs, last-mile delivery company, I recognized an opportunity to improve our inventory management process by developing an app that would help us make more informed decisions about where to stow inbound stock and manage replenishment.
By analyzing sales velocity, temperature requirements, and other key factors, the app would make recommendations on where to store new stock and suggest optimizations to our pick and pack layout, working in tandem with our location-based inventory system that maps all products to specific shelves in our warehouse.
Here’s how I used AppSheet to rapidly prototype and launch the app.
Step 1: Defining the Problem
The first step in building the app was to define the problem we were trying to solve. Our current inventory management app did not take into account data like sales velocity and storage requirements. Therefore the ops team often struggled to find the optimal locations to stow inbound stock within the warehouse location based inventory, making the inbound process quite time-consuming.
By extracting this key information from looker, and providing it to the warehouse team, we believed we could decrease time spent on put away while also optimizing the pick and pack route within the constraints of our location-based inventory system.
Step 2: Choosing the Right Platform
After evaluating several options, we decided to use AppSheet as our platform for building the app.
Key to this decision was how it facilitated us to quickly create prototypes and iterate based on feedback, its security and ability to connect to our data sources, including Looker and Google Sheets, and to build custom workflows and formulas.
Another important factor was maintainability, upkeep of the app had to be simple
Step 3: Designing the App With AppSheet
With AppSheet selected as our platform, we started to design the app. We collaborated with our operations team to create a user-friendly interface that would be easy to use, creating two views designed around each of the tasks operations team members may be performing, either stock inbound, or pick and pack optimization.
We also identified the need for a space planning view that displays a pie chart of how much of the inventory in percent is in each velocity-tier, category and sub-category, alongside the in-stock status of each slice.
Step 4: Building the App
Once the app was designed, we started building it in AppSheet and Google Sheets.
For pick and pack optimization, Google Sheets was used to pull location-based inventory data and then process and filter it into actionable lists.
For inbound, Google sheets processed a table of all product names, their bar-codes, storage requirements and current locations to determine their optimal storage location. Allowing the warehouse team to scan a bar-code into the search bar and find a products optimal location.
AppSheet was able to connect to our data sources quickly and easily, and we were impressed by the flexibility of the platform.
Step 5: Testing and Refining
With the app built, we started testing it with our operations team. Using multiple different types of device, Honeywell, phone and PC. We gathered feedback and made several refinements to the app based on the operations teams suggestions, data filters for inbound, swipe-to-mark-complete, for the pick and pack optimization list, the scan barcodes with a phone camera feature to name a few.
Because we had built the app in AppSheet, we were able to iterate quickly and easily, without needing to write much code.
Step 6: Launching and Training
Finally, we wrote an SOP (Standard Operating Procedure) containing a set of guidelines and best practices, to ensure that the app was being used effectively. Once this was complete, we launched the app and provided training to our operations team on how to use it.
Because AppSheet is so easy to use, our team was able to immediately able to get up to speed on the app’s functionality, including how it side-loads with our location-based inventory system.
Results
Thanks to AppSheet, we were able to rapidly prototype and launch our operations app within one week. By making more informed decisions about where to put inbound stock and being able to efficiently optimize the location-based inventory the ops team have successfully improved our pick times and improved our overall efficiency when stowing inbound stock. Our operations team has also reported that the app has been easy to use and has helped them make more informed decisions.
Conclusion
Using AppSheet to build our operations app was a game-changer. By leveraging its no-code platform we were able to quickly create a prototype and iterate based on feedback. We were also able to connect to our data sources quickly and easily, and to build custom workflows and formulas that were tailored to our specific needs. It also enabled us to deploy it at scale to the entire business domain. Moving forward, we plan to continue using App until these functions are integrated into GoPuffs own location based inventory App ‘Darwin’
Would you would like to take a look at the app yourself? Here are the links to a data masked version of the spreadsheet that provides data to the app, and a copy of the app front end that is connected to the sheet.
