The client organization is one of America's most beloved restaurants. A passionate man founded the beloved organization in 1968 to offer delicious, high-quality, affordable, sustainable, and responsibly sourced seafood to everyone. Today, the company has over 700 locations worldwide, and in 2019 it got involved in working to protect and preserve our oceans by mitigating plastic and fishing gear waste. At the end of 2020, the organization generated over $2.5 billion in revenue.
With a global presence, the organization's business processes must be effective and efficient to keep locations stocked and customers happy. These business processes generated high-volumes data that needed to be processed through multiple systems, a perfect mission for Robotic Process Automation (RPA).
The beloved restaurant chain selected JOLT as its RPA implementation partner to assess, implement, and maintain automation workflows that helped with data reporting regarding alcohol purchases and customer refunds. The goal was to minimize the amount of manual labor needed, the chance for human error, streamline data reporting and business processes, and help establish a Center of Excellence (CoE) to ensure the continuous scale of the automation program.
Optimizing alcohol purchase reporting
The first process that JOLT and the organization targeted for automation was the reporting of alcohol purchases. The company’s restaurants and vendors produced high volumes of data that needed to be extracted, compiled, and processed through multiple data systems to generate daily, weekly, monthly, & other required Accounting & Fiscal Calendar reports that reconciled data from disparate sources that resulted in a highly manual and time-consuming task for the organization’s staff. The reporting process was highly volatile and prone to errors due to month-to-month variations on the number of franchises inside the chain, as well as differences in data formatting between vendors. This variation required the automation to be flexible and robust.
By leveraging our Blended Delivery Model with UiPath’s RPA Platform, we provided the organization with project leadership, guidance, and assisted with the development, deployment, & implementation of multiple flexible, end-to-end automated solutions to support their restaurant’s tracking of Alcohol Purchase Orders into a web-application named “Harmony by iControl”. The Daily solution extracts and gathers Alcohol Purchase Orders from all of their restaurants each day per their Fiscal Calendar, processing over 180 items to create a daily report to track Purchase Orders, ensure supply to all restaurants, and identify any outstanding requests pending. At the end of each Fiscal Week, the Weekly solution reconciles the organization’s data against each restaurant’s sales data and vendor’s systems’ data to ensure no disparities exist, all amounts are accounted for, identify any un-fulfilled purchase orders or errors that need to-be remedied, then create the Fiscal Week Report containing upwards of ~1,800 completed and reconciled Alcohol Purchase Order records. Team JOLT experts also assisted with establishing and optimizing the company’s Center of Excellence (CoE) and RPA framework, allowing their operation to scale as the automation program continues to grow internally. Furthermore, we supplied their development team with foundational knowledge on best practices for RPA development, as well as starting points for code standards, re-usable components, framework overviews, & similar development and technical material.
The company was able to automate the daily, weekly, and monthly reporting of purchase orders and dramatically reduce the time required, improved the accuracy and availability of data, and eliminated common data errors and the vast majority of others. With respect to the Daily solution & report, prior to automation, these took around ~2 hours of manual effort to complete, however this has been reduced to a time less than ~5 minutes per day after deploying, ensuring all restaurants would have an optimal stock of alcohol supplies and reducing frequency of stock related issues in-store. Additionally, the weekly report generation, which previously took 5+ hours when performed manually, was also decreased to roughly ~5 minutes, ensuring the reconciliation of daily reports and all updates made across both organization’s and vendor’s systems, allowing the business users & Accounting Team to make decisions quicker and more efficiently. Moreover, before either automation was in place, not only could errors exist in the report’s data, but as a result of the reports requiring manual creation, the availability of them was impacted for all until the 1 to 2 individuals who create them do so. However, with the scheduled automations, both the Daily & Weekly Reports are generated on a set date & time, making it readily available for the necessary business users. Last, by leveraging our Blended Delivery Model to construct the automations, Team JOLT could enable the organization’s internal RPA resources to continue to develop, enhance, and maintain their automations without relying on our resources, which helped the organization to enhance the Automation Operating Model that Team JOLT instilled into a true Center of Excellence that now delivers RPA to the organization in a routine and repeatable methodology.
Better and faster refunds
The second automation JOLT developed for the organization focused on credit card transaction adjustments. This automation workflow aimed to automate enough work hours to offset the capital cost and establish an internal RPA resource team.
The processes inside the credit card transaction adjustments were time-consuming and problematic due to data inconsistencies. The amount of manual labor required to assemble the data before taking action was limiting analysts to act on the information prepared. The workflow's high volume and complexity drove up the forecast for additional full-time staff members needed to maintain this process.
Additionally, the process to issue credit card adjustments was problematic due to inconsistencies in the manager's adjustment requests, causing additional work to research and identify critical data elements.
What JOLT decided to do first was to redesign the existing solution to correct the data via automation. In this automation, the requests were downloaded from the request portal and passed to the data warehouse to compare and include accurate and additional data points to process the adjustments automatically.
JOLT also determined that any request that did not return complete details was to be marked as an exception for a human staff member to analyze.
Once the robot returned the data from the data warehouse, it appended the data to the requests. Items that robots had already processed as an adjustment were exported to the accounting system for automatic journaling. Subsequently, the records that the system had not processed for adjustment were processed by the robots at the appropriate card site to issue the customer's refunds when a defined criterion is met. Furthermore, all adjustments made are updated in the request record and marked for exporting to the accounting system.
We are receiving great feedback on the credit card adjustments' time savings when the bot completes all this effort off hours. We arrive to review and validate then handle the exceptions which allow us to look for irregular activity and improve the managers' overall request accuracy. The team would have had additional overtime to catch up, and when we see any spikes in activity, we know we will see most of the information successfully processed by the bot and not impact the workload of the team going forward. What a person was doing in 4-6 hours is completed in a couple of hours off-hours by the bot and freeing up time for additional activities and analysis without growing the team.
By having a robot download the requests, analyze the data with the data warehouse, and automating the accounting extract, staff members at this beloved organization were saving 1 to 2 hours of manual labor every day. Additionally, by leveraging RPA to validate the data against the data warehouse and providing additional details, JOLT helped improve staff members' efficiency to be more productive at working out exceptions and adjustments that are not included in the automations scope.
Automating the primary site where over 65% of daily requests are received has saved the organization an additional estimated 4 to 6 hours of manual labor per weekday. All of this processing is performed off-hours, so the organization's team arrives each day to manage only exception data that the bot could not process.
The results can be summed up to the improved efficiency of the manual process which went from 1 day of manual work to 1 hour of bot work instead, eliminating 8 hours of manual work daily, saving up to 40 man hours weekly. The staff was freed from tedious manual work and gained the ability to spend their time on higher value tasks.