Address
268 Bath Road
Slough, SL1 4DX
Phone
+44 (0)1753 708 529
Overview
In the fast-paced world of FMCG, quick access to accurate data is critical for making strategic decisions. But what happens when the data you’ve been relying on turns out to be inaccurate?
This is a case study about a manufacturer who had no reason to assume their monthly sales data was anything but correct. Reports were generated and ready to be distributed to key stakeholders throughout the organisation, including senior leadership teams, who were eagerly awaiting the latest sales trends.
The Challenge
Some troubling patterns that were difficult to detect were present in the data, these appeared to stem from restatements and miscoded SKUs. The data also showed key brand sales declining when growth was expected. This was coupled with no communication to the manufacturer from the data provider regarding expected anomalies or restatements in the most recent month’s data update.
Examination of the data exposed two issues:
- Incorrectly reported volume sales and price of the manufacturer’s newly launched products due to miscoded pack sizes.
- Numerous miscoded SKUs were placed in the wrong category. This inflated total category sales, and depressed market share for this manufacturer – reporting value share decline equating to a $70 million mistake in one monthly period (-3.5pp), when in actual fact share had risen.
The Solution
Wessex Data Quality Control protocols are embedded into the manufacturers processes – with the use of AI machine learning models, abnormal changes to the data were revealed.
AI Machine Learning
QC
Dashboard
Error
Report
24 Hr Turnaround
Wessex/Agency Collaboration
Errors Highlighted
Deep-dive investigation
Report triggered
Error report to data agency
Corrected data
Possible disasters averted:
- Incorrect data containing $70 million mistake reported throughout the business.
- Lengthy, resource and time-consuming investigations – this means time, effort, resource, and money.
- Incorrect remedial actions – diverting marketing spend due to the poor performance indicated.
- Incorrectly reported price negatively impacting relationship between the manufacture and its retailers.
- Erosion of trust.
Positive outcomes:
- Key business performance indicators continued to be correctly reported throughout the business.
- Strategic and commercial decisions were made on robust, high-quality, accurate data.
- Confidence, trust, and engagement with the data was maintained.
- Value share growth was correctly communicated as a success.
