ClusterTruck is poised to turn the restaurant industry on its head. Instead of a walk-in restaurant with a permanent location and wait staff, they operate delivery-only restaurants. Their engineering team developed a suite of web and native applications to drive and manage their entire business, from the ordering website to kitchen management to driver apps.

No visibility into a complex process across apps and stacks

Unlike most web applications, ClusterTruck outputs hot food to your door, not HTML to your browser. Food is ordered through a web application, cooked and packaged in a kitchen managed by custom Android appliances, Clustertruck app and delivered by adriver using a smartphone and his own car (or bicycle!).

Complex routing rules guide the entire process. For example, orders aren't sent to the kitchen application until a driver is available to deliver the food, and burgers must be on the grill three minutes before starting the pizza.

Some developers may have the luxury of believing they only need to monitor basic server stats, but not Dan McFadden, ClusterTruck's Co-Founder and CTO. Dan needed visibility into the entire process - from start to finish - in real-time. Behind every order was a hungry person, awaiting fresh food, who will immediately notice any mistake or delay.

Monitoring software at the code level and up

Dan chose Instrumental because it provided ClusterTruck with the flexibility and power to monitor every part of their infrastructure. From database servers to iOS apps, Instrumental measures the status of each component and the success of every order. This is only possible because ClusterTruck is monitoring their software at the code level. In a sense, Instrumental serves as a real-time, production-level test suite.

Finding opportunities and fixing problems faster

With complete visibility into ClusterTruck's platform, Dan and his team now quickly debug any issues that arise. They also proactively look for performance improvements. Using Instrumental, ClusterTruck determines what percentage of drivers return after an order. Clustertruck graph Their updated algorithm accounts for the number of drivers returning shortly, so they begin cooking orders sooner.

This results in an improvement of more than 33% in driver wait times, and an accompanying decrease in customer wait times, all visible and verifiable with Instrumental. ClusterTruck made further improvements by sharing driver-demand Instrumental graphs inside their driver mobile app, so drivers identify when delivery capacity is in high demand.

Monitoring for marketing and operations

After walking by the engineering team's wall-mounted dashboard, the ClusterTruck marketing team asked for their own Instrumental dashboard. Clustertruck dashboard Marketing emails and social media campaigns are now triggered by real-time stats on how orders compare to yesterday's or last week's orders. If group orders are down today, ClusterTruck emails previous group-orderers a special offer. If the pizza station in the kitchen is overwhelmed, ClusterTruck emails users to remind them ClusterTruck also serves tacos. By leveraging real-time data across their entire application, ClusterTruck operations react quickly to market changes and deliver a better experience, while maximizing use of available resources.

Questions? We can help!