AIOPSMonitoring™

After deploying the monitoring solution, AIOPSGROUP identified that there is an intermittent issue with PayPal that was leading to failed sessions and frustrated customers. After further investigation, the Incident Report concluded the impact of this issue was 120,000 EUR on average of lost revenue per month. The customer confirmed a 2.5 million EUR loss in revenue over time by not fixing this issue.

Post initial implementation, the AIOPS Monitoring team noticed that there are as high as 374 attempts on a single order and a lot of fraudulent attempts during peak times, testing all available payment methods. Although the customer had a robust fraud prevention system in place, in 1% of cases end users were able to place orders. Once the issue was fixed, this resulted in more than 2.5 million USD saved in charged-back fees. In addition, AIOPSGROUP created a fraud detection dashboard where the customer’s team has a list of all suspicious orders and was able to investigate and cancel fraudulent orders.

The payment instrument, Klarna, disappeared from one of the customer’s high-traffic websites. This was identified 5 days later. The iframe was simply hidden by security, so no errors were logged on the front end or back end. AIOPSGROUP implemented Payment Velocity monitoring, based on the last 30 days. The alert is triggered even if a certain payment instrument stops performing on a single website, maximizing the revenue stream of the orders already in the system.

A customer had several thousand orders stuck in an intermediate state and not processed to OMS. If this had continued, these orders would have been canceled. AIOPSGROUP monitoring reported the issue in a matter of minutes preventing the orders and revenue in the system from failing.

Over the last 15 years and 250+ SFCC implementations, AIOPSGROUP observed the wrong or delta file uploads, where major products and availability are overwritten from the customer side. AIOPSGROUP is monitoring for changes in product availability, online products, products without images, products online without availability, products without proper taxation setup, and raising alerts for inconsistencies.

AIOPS Monitoring identified a high rate of failed customer sessions (15-20%). After root cause analysis, AIOPSGROUP managed to link sessions to specific template rendering errors coming from the SFCC logs. Resolving this issue resulted in additional revenue of 1 million EUR per year.

Due to an export issue, several of the customer’s products were set at the price of 0 on the storefront. Consequently, this issue managed to generate around 800 orders of 0 value or with products containing products with a value of 0. The losses in revenue were significant as all those orders needed to be canceled. AIOPSGROUP set up an alert to monitor such scenarios, in addition eliminating this possible issue for all the existing customers.

Per Customer’s request, AIOPSGROUP improved the solution to allow for ingestion of SFCC OCAPI data and logs in a frequency of 1 minute. This allows the Customer team to have a nearly real-time view of the technical performance of the sites and be able to allocate resources if an issue arises, increasing the total uptime of the sites.

AIOPS Monitoring recorded random Payment Provider’s 4xx response code rather than success. This was confirmed but root cause analysis was not able to determine the cause and timing. This was not a bug in the code but a Payment Provider failure. Retry logic was implemented to retry the Payment Provider call up to 3 times before failing the order.

A sports brand made an internal error and placed all their products at 60% sales. This customer triggered a large online sale. As a result, AIOPSGROUP has implemented price change KPI monitoring in such scenarios.

There are a lot of customers with 20+ different websites and multiple jobs per site. Whenever a job is stuck it cannot be monitored manually for all sites. AIOPSGROUP implemented KPIs to monitor Patterns in Job errors visible in the dashboard. There are alerts triggered and escalated for failed jobs.