6 Most Painful Mistakes in Spend Analysis

A guest blog by Sammeli Sammalkorpi, Co-founder, Sievo.

While procurement leaders increasingly turn spend data into a competitive advantage laggards continue to repeat painful mistakes.

You may have noticed spend analytics is a hot space right now. According to Deloitte, the majority of CPOs across the world expect analytics to have the greatest impact on procurement over the next two years.[i] As one of the leading spend analytics software providers we’ve also been busy. At Sievo, we’ve seen more interest than ever from both private and public organizations in getting their spend data in shape.

As leading procurement organizations find new opportunities to turn spend data into value, we see others repeat painful mistakes. Here are the top six mistakes you should avoid in your spend analysis projects.

1. You can’t manage what you can’t measure

We still come across far too many organizations that don’t have full visibility on their procurement spend. Some may have limitations from IT on the software they can use and some may lack the skills in-house to combine spend from different source systems. Having many ERPs and source systems is no excuse, you can’t manage what you can’t measure.

2. Bad data is no better than no data

Another common mistake we see is that smart procurement professionals need to rely on unreliable spend data. The chances for mistake are especially high if your still conduct your spend analysis in Excel or infrequently done manual spend cubes. Rarely can you make good decisions based on bad data.

3. Don’t rely on the United Nations to classify your data

You may have heard of UNSPSC, the taxonomy standard developed by the United Nations and still commonly used to classify products and services in procurement. While a universal standard taxonomy is an admirable vision, most complex organizations have a far better understanding of their own spend. You know your business better than the UN.

4. Analytics is not just about fancy graphs and dashboards

Thanks to the wide adoption of business intelligence (BI) tools, analytics has become widely engrained across most organizations. Still, it makes me shiver when I hear the question “can’t we just do this in Tableau or Power BI?” Analytics is much more than graphs and dashboards. In spend analysis, 90% of the challenge comes from getting the right data and insights in view. Once you’ve mastered the data, the graphs are just window dressing.

5. Don’t just look back at past spend data

For too many organizations, spend analysis is a reactive task, only looking at past spend performance. Leading organizations take a forward-looking view on spend, for example through procurement spend forecasting or proactively benchmarking procurement performance  with peers or market indices.

6. Letting humans to do what machines can do better

At Sievo, we’ve made more advances in Machine Learning spend classification in the last twelve months than we’ve made in the previous fourteen years in manual spend classification. We still believe that experienced procurement professionals need to be at the helm of spend classification, but we see enormous potential teaching machines to take care of repetitive, complex and time-consuming processes.

That’s enough of dwelling on mistakes. If you’d like to hear about things you could do right in spend analysis, reach out to your friends in Cloudia for an introduction to Sievo’s spend analysis solutions. We’ve always got time for friends of our friends, and you may find we’ve got even more in common!

Author Bio

Sammeli Sammalkorpi is the co-founder of Sievo. His top focus is to ensure Sievo’s customers can turn their procurement data into realized business value. Originally from Lapland, Sammeli is a father of three, youth football coach and avid traveler.

[i] 2018 Deloitte Global CPO Survey

Sammeli Sammalkorpi,

Co-founder, Sievo