Optimizing audit selection 

Legacy audit selection relies heavily on auditor-selected leads—a process that can be labor-intensive, inefficient, inaccurate, and prone to bias. The task of selecting audit leads is far better suited for algorithms that can comb through large amounts of data in the blink of an eye. Voyatek’s RevHub Audit Optimizer leverages NLP and ML to help states identify and prioritize quality cases for audit while minimizing manual screening and scoping processes. For example:

1.  Clustering algorithms can detect the most likely non-compliance for sales and use tax.
2.  Ensemble modeling identifies artifacts associated with non-compliant corporate filers.
3.  Machine learning can find events associated with individual underreporting.
4.  Complex correlation analysis can uncover non-compliance for pass-through entities.
5.  Zapper data mining offers big data analytics for point-of-sale transactions. 

Altogether, these approaches generate quality, data-driven audit leads that minimize auditor bias while bypassing the inadequacy of manual audit techniques. RevHub’s Audit Optimizer can ingest data from any source and ensure that every return gets analyzed. It can be deployed as an Integrated Tax System (ITS) or as a point solution that’s integrated with an existing ITS. For the cherry on top, the solution quickly and easily adapts to policy changes. 

Missouri’s success story 

Already, customers are enjoying the benefits of RevHub’s Audit Optimizer. Agencies across the country have increased audit ROI by 31%, decreased audit time by 20%, and brought in more than $105 million in assessment funds using our tool. The Missouri Department of Revenue’s Field Compliance Bureau (FCB) is one such success story. 

In a nutshell, Missouri’s Department of Revenue attempted to modernize its audit system, but the predictive audit model it implemented did not live up to expectations. The new ITS was supposed to be paid for by the increased revenue it was able to bring in, adding another layer of disappointment.  

FCB urgently needed a solution capable of increasing audit success rates, while maximizing resources. So, Voyatek created an expedited schedule for layering RevHub’s Audit Optimizer solution on top of FCB’s existing data warehouse and business intelligence platforms. 

Audit Optimizer in action

A key component of RevHub’s Audit Optimizer solution is its Audit Selection Models, which leverage machine learning to score entities and their sales tax filings. Then, the models output rank-order audit candidates based on statistical deviance from population norms.  

This is possible because the solution curates internal and external data sources, including account, entity, and filing information for all tax types from the ITS; federal corporate and 1099K filings; state Department of Labor and wage data; lottery sales; audit history and case management information.

Finally, an extensive Extract, Transform, and Load process maps that data into a 360-degree view of taxpaying entities and their filing histories to facilitate audit investigations. The results have been undeniable. Since implementation, Missouri has used RevHub’s Audit Optimizer to:

–   Analyze more than 750,000 entities across industries for non-compliance issues.
–   Yield more than $108 million in audit assessments.
–   Increase audit profitability return on investment (ROI) by +25% findings/hour.
–   Reduce average “time to complete” metrics by 18%.

The bottom line is that many state agencies are facing budget cuts, while tax revenue growth is expected to cool in the fiscal years ahead. Thus, the time is now for tax agencies to implement cutting-edge technology to maximize both short-term revenue and voluntary compliance. 

To learn more about how RevHub’s Audit Optimizer can support your agency’s audit efforts, get in touch today.