The first thing we did was a quick assessment and analysis to figure out the root causes behind the instability and the lack of accuracy of the application systems.
We re-factored the most critical integrations and added error handling and alerting and then brought the web servers up-to-date
The impact of our work with this platform is clear: when we started, the company had around $40 million in revenue. Their revenue is a little over $100 million now.
Our client is a traditional retailer with a growing eCommerce platform. They put together this eCommerce architecture in an extremely rushed fashion, and they were struggling with the stability of the application, and the accuracy of the application.
They approached us because our values and model were extremely interesting to them and because they had a few stores on the West Coast and a few stores on the East Coast which they were not supporting adequately. They wanted to make sure that there was a reliable team that could proactively ensure availability and performance and that could extend the coverage of the support system.
The first thing we did was a quick assessment and analysis to figure out the root cause behind the instability and the lack of accuracy of the application systems. Many times the inventory numbers would update incorrectly. We did a very quick, basic, six week blocking and tackling job. We figured out which were the most critical integrations. We re-factored those integrations and added error handling and alerting capabilities, and then brought the web servers up-to-date with the necessary software upgrades and patches to reduce the repeated failures. In six weeks’ time, the number of incidents was reduced by almost 60 percent. Then, as we took control of the application and integration systems, we continued to not only ensure the availability and stability of the application, but we also kept refactoring a lot of “spaghetti code” on the integration layer, as well as inside the application.
Before we arrived, their approach to solving the integration errors was to try to manage the incidents. What we hoped to accomplish with “basic blocking and tackling” is that, unlike managing the incident, we would take care of the root cause, rather than just enacting problem management. Blocking and tackling is fully solves the problem so incidents don’t repeat. We were able to show, month over month, quarter over quarter, and year over year, a consistent increase in system availability with a reduction in incidence rates. We put the application on a more stable path and the business comes to us with new ideas to enhance system features. When developing an e-commerce engine and website system, skilled, outsourced external teams can optimize and integrate these systems with other business IT assets.