Cox built an asset tracking platform that allows users to attach tracking devices to their important assets. Users need to associate assets with a specific tracking device, allowing them to later find the asset in the dashboard.
In the first weeks after founding, we created a simple pairing app. I no longer have the designs for it – it was all done on pen and paper and shipped within a week. We affectionately named this version "Boop de Boop" after the sound made by the hand scanners used in testing. Evenutally we opted to use cell phone cameras over hand scanners since most of our customers issued some sort of smart phone to employees. Eventually our customers began to out grow the simplicity of early versions. When talking with them, they expressed several concerns:
Warehouse users can easily pair hundreds of assets a day. Every click and screen greatly impacts their workflow. Although warehouse employees responsible for pairing weren't our biggest champions, they were usually the loudest when things stopped working.
Our early customers in the medical industry used standardized data (SKUs, Serial Numbers, etc.) with GS1 barcodes. This made initial implementation extremely easy since the inputs were predictable. As we expanded into manufacturing environments this was no longer true, so we needed to be more generous in what we accepted.
User's also complained that finding assets in the dashboard was growing more complicated. We found this was directly tied to the growing data complexity and inconsistent naming of assets. While visiting one of our customers in Detroit, I found users were having difficulty understanding what to name an asset. Since they were not responsible for finding assets in the dashboard and their metrics were tied to processing shipments quickly, they would make their best guess and quickly move on.
We can do several things to streamline the workflow when pairing/unpairing an asset
Since the users pairing trackers are not the end user of the dashboard product, they do not always understand what data is important to capture. We can solve this by allowing admin users to designate required and optional data fields. Additionally, we can enforce certain formatting standards for specific data. The new app also can scan multiple barcodes in one go and automatically map them to a customer's data schema.
We discovered many of our users have planning documents or BOMs that outline what assets they plan to consume. We can ingest these documents to enable autocomplete for manual entry. This also allowed us to enable suggest names, where we derive an asset name based on the inputted data. Inconsistent asset naming was the number one complaint of project managers using the platform.
We greatly simplified the workflow and number of screens required to pair an asset. If there are no errors, a user only sees three screens and clicks one button to successfully pair an asset.
There are edge cases where it is not possible to scan barcodes. We implemented a manual entry mode that allows users to bypass scanning.
Based on feedback from early releases, we shipped a few quality of life enhancements that catch problems earlier in the pairing process.