Tail spend: Use Case - Unexpected Insights & Takeaways

There are many articles written on tail spend, on the importance of visibility from a compliance and risk perspective as well as on the value that procurement teams leave untouched. We agree with all the above and want to share our takeaways from a recent project where we refined the organization’s owned spend taxonomy and brought visibility into the full range of previously unclassified spend and had the procurement team shocked when looking under the hood”.

Use Case

A multinational company with over 30 manufacturing locations across the globe, 20+ source systems (ERPs) and over 100.000 part numbers (only on direct spend), has established its own spend taxonomy consisting of 4x levels and initiated the allocation of those (manually) into their key supplier’s products and services a few years back.

However, the outcome only represents less than 1/3 of the suppliers and their associated spend. The procurement team has worked for over 3 years to maintain current scope and hardly managed to add anything significant during these years due to time and resources constraints, as well as to ”lack of energy and passion to do the endless number crunching” according to their CPO.

The general feedback was that the remaining of the spend is tail spend only and therefore it isn’t worth the time and effort to gain visibility. However, leaders of the organization were convinced given the amount and volume of unclassified spend, there is untapped potential to be gained and were looking for support to explore it.

Mission

Our mission was to capture all spend in one place without introducing yet another new IT project, and to help the team to look into what they’ve developed through a different lens to refine and simplify it. Last but not least, to create full visibility into their entire scope of spend using their own spend taxonomy.

Using Mithra Ai’s drag & drop feature, procurement teams were able to add all ERP extracts into our cloud and generate a holistic spend view in a matter of days. As the next step, the raw data has been parsed and cleansed automatically and reviewed by exception when needed. This is a key enabler for the classification step as well as insights which many teams and/or organizations typically avoid doing. In parallel, reviewing and refining current taxonomy was a must exercise. We call this step the “taxonomy health check“.

The procurement team received a list of suggested imperfections found by our machine learning algorithms on their homegrown spend taxonomy. Their spend taxonomy was developed with 4x levels and 434 unique branches making it a complex taxonomy tree, even on the direct side. Once the team reviewed flagged items and applied the suggestions, the final and improved taxonomy was locked as the main scope for this activity.

Finally, for spend classification, a series of hundreds of machine learning models have been designed and assigned to classify all spend within the scope (with overall confidence level of 98%). As the last step, the proposed classifications have been reviewed (by exception) by each category team and business rules have been applied when needed. Hard to believe for many but the above exercise took only 3 weeks.

Result

Now the procurement team had access to an untapped ocean of first time ever classified spend without doing the heavy lifting in less than a month. But the outcome was nothing close to initial expectations: spend with transactional suppliers was actually quite low.  

Across all suppliers within uncategorized spend, over 28% of them were “strategic or critical” supplier segments which represented over 80% of the total uncategorized spend.

This was completely unexpected and came as a shock for the procurement team. Their first set of actions was to go back to those suppliers and review terms and contracts. A good portion of this additional spend was either services that never have been part of the original taxonomy (so the team simply ’t know the company was buying these products, services) or new businesses given to these suppliers without categorizing those (looping category teams). Now, the category team has over $300 mln additional in their pipeline of opportunities and is fully focused to work cross-functionally and realize this value both at the top and bottom line for the organization. This makes this group heros especially in this market dynamics and they know they have a good engine that takes care of the maintenance of their spend pipeline.

Takeaways

1. You will never know what value is under the hood (your tail spend) till you have visibility to it
2. Taxonomy development and maintenance is key – otherwise you’ll be shooting in the dark
3. Let the right technology work for you instead of pushing tedious work to your teams and make them disengaged with the purpose – Empower the team to focus on high value activities instead.

Last but not least – have a chat with us and let us inspire you!

‍