The cube consisted of hundreds of thousands of rows on customer-product level spanning across several business units – this meant that prioritisation of the most impactful cases was key. To find and correct the most impactful cases we took the following three-step approach:

• Define the prioritisation metric: The core aim of the project was to reduce incorrect downsell and churn. Therefore, we prioritised customers by their total ARR loss (churn and downsell) over the period of interest.
• Bottom-up approach: We analysed individual customer’s ARR profiles and used supporting evidence to create adjustments to appropriately distribute ARR across the time period of the invoice, removing false ARR loss at the customer level.
• Top-down approach: We then explored customer segments with high levels of ARR loss to identify systematic business logic errors and rectify them across the whole customer segment, eradicating large clusters of false ARR loss.

This methodology allowed us to create a list of suggested adjustments and customer re-mappings to implement, focussing on the first two challenges the client outlined to us.

Auditability was also a key success criteria. To achieve this, we worked closely with the client through a series of workshops where adjustments would be reviewed and accepted. Supporting documentation was prepared through this to justify adjustments and remapping.

To overcome these challenges, we built a self-serve data platform for the client which would support our client in increasing their planning horizon. We leveraged our data capability and reporting experience by focussing on the following activities:

1. Understanding the Client’s Needs
   We began with engaging the C-suite, finance, and sales operational leads to understand their pain points. These pain points were collated and used to design the reporting tool.
2. Data Assessment and Process Recommendations
   Our team evaluated the level of detail available within the client’s existing data systems to perform a feasibility assessment of the reporting tool design. Where data was unavailable, process improvements were outlined to enable more granular reporting in the future.
3. Reconciliation and Data Quality Improvement
   A rigorous reconciliation process resulted in the creation of a clean customer ARR cube. This process corrected historic manual errors, particularly in invoice dates, and significantly improved data quality.