Domino's Pizza
Modern Store Locator
Did you know a "gps store locator" algorithm can be patented? I didn't - but that's what Domino's was dealing with in the mid 2000's as online ordering became more prevalent. The easiest way to locate a store relative to a customer, using geofencing and geolocation, was a patented process. As a result, they were forced to build a more complicated, and harder to maintain system.
Approach
With that starting point, I was tasked with helping create a new Modern Store Locator. This was an AKS application designed to use a customers' geographic location to find them the local franchise designated for delivering their pizzas - or the closest possible franchise for them to order carry out (though these are usually the same)
This meant answering a lot of questions:
- How in the world does the legacy version work?
- Are we responsible for web clients or native app clients, or both?
- What about 3rd party clients like Uber Eats?
- Who decides what areas each franchise is responsible for delivering to?
- How is that data created? Where does it live?
- How can we determine the geographic location of a customer?
- *Can* we determine their location? Do we need a data partner?
- How can we relate the customer location to the data deciding which franchises are responsible for?
- How can we keep these data points up to date in real time?
- How can we keep this flexible in case franchises decide to open or close, or trade territory?
- And many...MANY more
Answering these were the foundation of my work, and staying close with architecture to decide the best means of systematically answering each one are how I built our backlog to get an MVP of the new service up and running. The new service is more accurate than the previous one, less prone to data errors, faster to update, and fits into a quicker SDLC. Overall a successful project, executed over the course of autumn 2024.
Key Learnings:
- First time myself or my team had worked with AKS and AKS deployments.
- Getting buy in from stakeholders entrenched in the legacy system isn't ever easy - even with senior leadership alignment it can take some time building relationships and demonstrating future value to get the full ship moving in the right direction.
- This was my first consumer facing application in 7 years, I almost forgot how fun it is to play with your new application online and immediately know that thousands of real paying customers are using it at the same time.
- Being the guinea pig for a new SDLC isn't very fun - lots of "learning together" moments when you really just want to ship your Proof of Concept and start iterating on it.
- Geographic data is an enormous beast unto itself - I had no idea the geographic information systems industry is a $20 Billion dollar industry of it's own right.
- My first foray into a Google API - some had told me it would be a nightmare but I found it fairly straightforward. The billing and metrics interface could definitely use some TLC.