Written by Matt Holzapfel
What if every credit card statement you received had a 25% charge for “other fees”? Or what if a friend asked you for $2,000 so they could buy “other?”
One of the best parts of my role at Tamr is using cutting-edge technology to show customers where all their “other” spend is going. This has led to cost savings of tens of millions of dollars for some of our customers. But perhaps even more importantly, it empowers them to take control of their spend now that they finally have a complete picture.
There are three primary reasons unclassified and misclassified spend is such an expensive problem for customers.
Non-Compliant Spend Gets Hidden
There are many forms of non-compliant spend. One of the least concerning forms occurs when an employee is acting in good faith, but has a preferred supplier that is different from the procurement-designated preferred supplier (i.e., off-contract spend). The more expensive forms of non-compliant spend happen when an employee is intentionally spending money in a way that is damaging to the company, such as fraudulent spend or spend that circumnavigates internal budgets.
The total value of non-compliant spend is significant. Every dollar of off-contract spend results in a loss of 12 – 18% to the enterprise. Spend that is fraudulent or would not have been approved is a loss of 100%. A company with $1 billion of non-compliant spend could easily be spending $150 million unnecessarily.
Reducing non-compliant spend is only possible when it can be easily recognized by procurement teams – a difficult task when large amounts of spend get buried in “other” or the wrong category in spend reports.
Sourcing Strategies Become Based in Fiction
Spend must be properly understood in order for procurement leaders to effectively design their organization and sourcing managers to execute. Within indirect spend, lacking detailed spend information makes it difficult to consolidate suppliers and identify opportunities to bring new items or categories under contract. Within direct spend, poor categorization of spend makes it difficult for procurement to recommend alternative suppliers and parts to engineering, hindering their ability to add value beyond price negotiations.
The net result is that sourcing strategies become based on an assumed understanding of spend, and not actual data. This is similar to driving to a new city without a map or GPS, hoping there are enough signs along the way to point you in the right direction.
Executives and Managers Stop Trusting the Data
The CFO is consistently recognized by CPOs as one of their two most important C-level relationships. This makes it essential that CPOs be able to communicate in terms of data and numbers that are accurate, trustworthy, and detailed. No CFO wants to hear that costs are above forecasts because of “other”.
The most obvious impact of bad data is that sourcing teams will spend significant time scrambling to put together data to understand the root cause of issues. Less obvious, but more important, is that senior executives will lose trust in the recommendations the team is making – knowing that they are operating with incomplete or inaccurate data.
The customers we’ve helped solved this problem have seen transformations in how data is used. Executives stop using meetings to grill people about the accuracy of the data, focusing instead on how to solve the business problem. Managers feel more confident presenting insights, knowing they can trust what the data shows. This helps create a culture of transparency and trust, while accelerating decision making processes.
Saying Goodbye to “Other”
The misclassification of spend data has been a persistent problem for years. A sustainable solution to this problem has finally become possible with advancements in machine learning technology from products such as Tamr. Machine learning is essential to solving this problem because of the dynamic nature of purchasing – the items an organization purchases and how buyers describe these items are constantly changing. Adapting to these changes to maintain trustworthy data requires machine learning that can be trained by its users as these changes occur.
Your company can finally stop paying a tax for “other.”
To learn more read the white paper.