Ask a simple question – what is clean data?
Some of the answers:
“That’s a hard question to answer.”
“It means you can have 100% trust in its accuracy.”
“Speaking with procurement teams from all industries and sizes, we know that data quality is still a real challenge–they’re struggling with disparate systems, manual data entry processes, and they aren’t able to deliver all of the value to the organization that they would hope to. Organizations want to fix their supplier data, but they don’t know where to start.”
A Dewey Decimal Classification System For Procurement?
As I pondered the above answers, the first thing that came to my mind was that they were talking about the Dewey Decimal System or the Library of Congress Classification System. The only difference is that the information being cleansed or managed wasn’t about books but suppliers.
In a way, these new systems for cleansing are really managing a static environment rather than delivering actionable knowledge that can improve decision-making regarding crucial performance metrics. The only difference between books and suppliers is that the text of The Merchant of Venice won’t change, but a supplier’s certifications or classification status can. While necessary for compliance, how much does this passive management of data actually lead to better decision-making beyond said compliance?
What Is Real “Data Trust”
Many years ago, a major U.S. retailer flew me in for an objective third-party assessment of their organization’s vendor rationalization program.
Two years prior, they had decided to reduce the number of suppliers, concentrating all of their business on a select 100 suppliers, with the goal of lowering supplier administrative costs and increasing volume discounting.
At first, the strategy seemed to be working because they realized immediate savings to their bottom line and considerably less administration with even less paperwork. In short, money and FTE cycle times decreased considerably.
So, why call me in to do an assessment? To start, the earlier level of savings was not as robust as they had expected.
Using a self-learning algorithm-based solution, which worked so well for the DND, I expanded their normal buys through the preferred 100 to include a few hundred suppliers with whom they were not dealing.
Redefining “Data Truth”
Let’s jump to the results of my expansion of supplier engagement beyond their preferred 100. This large retailer was paying, on average, more than 20% above the going market rate—twenty percent!
The issue was that they were too focused on compressing the number of suppliers to a more manageable number at the expense of engaging a more significant, ever-changing open market.
However, the following demonstrates the main difference between the static management of existing supplier information and dynamic response to a volatile market.
With every transaction, the self-learning algorithms would quantify historical supplier performance in critical areas like past delivery performance, product accuracy and quality, and current-day performance in geographic location and time-of-day impact on pricing, to name a few. By the way, the buyer could adjust the weighted value of each of these performance areas. The ability to adjust weighted importance at the buyer level meant that if next-day delivery was the priority, the algorithms would calculate which supplier had the best score to meet that need. Conversely, if pricing was more important than delivery time, changing the weighting of the system would automatically, within seconds, re-rank the suppliers.
The result was that the algorithms of this nascent AI platform would recommend the right supplier 97.3% of the time. The more the platform was used and expanded, the “smarter” and more streamlined it became.
You can use this link to find out more about the platform and specific results – https://bit.ly/3tj9qhB
In the meantime, what does “Data Truth” mean to you? More importantly, what do you define as “actionable” knowledge or insights?
30
Procurement Insights
What is “clean data” – how do you define it?
Posted on July 10, 2024
0
Ask a simple question – what is clean data?
Some of the answers:
“That’s a hard question to answer.”
“It means you can have 100% trust in its accuracy.”
“Speaking with procurement teams from all industries and sizes, we know that data quality is still a real challenge–they’re struggling with disparate systems, manual data entry processes, and they aren’t able to deliver all of the value to the organization that they would hope to. Organizations want to fix their supplier data, but they don’t know where to start.”
A Dewey Decimal Classification System For Procurement?
As I pondered the above answers, the first thing that came to my mind was that they were talking about the Dewey Decimal System or the Library of Congress Classification System. The only difference is that the information being cleansed or managed wasn’t about books but suppliers.
In a way, these new systems for cleansing are really managing a static environment rather than delivering actionable knowledge that can improve decision-making regarding crucial performance metrics. The only difference between books and suppliers is that the text of The Merchant of Venice won’t change, but a supplier’s certifications or classification status can. While necessary for compliance, how much does this passive management of data actually lead to better decision-making beyond said compliance?
What Is Real “Data Trust”
Many years ago, a major U.S. retailer flew me in for an objective third-party assessment of their organization’s vendor rationalization program.
Two years prior, they had decided to reduce the number of suppliers, concentrating all of their business on a select 100 suppliers, with the goal of lowering supplier administrative costs and increasing volume discounting.
At first, the strategy seemed to be working because they realized immediate savings to their bottom line and considerably less administration with even less paperwork. In short, money and FTE cycle times decreased considerably.
So, why call me in to do an assessment? To start, the earlier level of savings was not as robust as they had expected.
Using a self-learning algorithm-based solution, which worked so well for the DND, I expanded their normal buys through the preferred 100 to include a few hundred suppliers with whom they were not dealing.
Redefining “Data Truth”
Let’s jump to the results of my expansion of supplier engagement beyond their preferred 100. This large retailer was paying, on average, more than 20% above the going market rate—twenty percent!
The issue was that they were too focused on compressing the number of suppliers to a more manageable number at the expense of engaging a more significant, ever-changing open market.
However, the following demonstrates the main difference between the static management of existing supplier information and dynamic response to a volatile market.
With every transaction, the self-learning algorithms would quantify historical supplier performance in critical areas like past delivery performance, product accuracy and quality, and current-day performance in geographic location and time-of-day impact on pricing, to name a few. By the way, the buyer could adjust the weighted value of each of these performance areas. The ability to adjust weighted importance at the buyer level meant that if next-day delivery was the priority, the algorithms would calculate which supplier had the best score to meet that need. Conversely, if pricing was more important than delivery time, changing the weighting of the system would automatically, within seconds, re-rank the suppliers.
The result was that the algorithms of this nascent AI platform would recommend the right supplier 97.3% of the time. The more the platform was used and expanded, the “smarter” and more streamlined it became.
You can use this link to find out more about the platform and specific results – https://bit.ly/3tj9qhB
In the meantime, what does “Data Truth” mean to you? More importantly, what do you define as “actionable” knowledge or insights?
30
Share this:
Related