Customer Stories

Crunch time for AI-powered supplier discovery

Written by Admin | Apr 11, 2025 2:39:48 AM

FMCG

Finding new suppliers isn’t easy. Especially when the desired product is hard to come by - as was the case with frying oil for a potato chip company that needed to launch a new type of potato chip with a million-dollar revenue plan seemingly paused.  

It had been three weeks without results. No matter where they looked, the procurement innovation leader and one of his team members hadn’t found a suitable supplier. Nothing seemed to be working - weeks of on and off searching, including calling up their network of known suppliers, repeated Google searches, trying every supplier repository they could think of, and sorting out duplicates - all it had done was cost them a lot of time and energy.

Reporting this back to the R&D team would call a halt to the plan they had meticulously developed. Months of customer research had shown that there would be significant demand for this new type of potato chip. Canceling this project would reduce a million-dollar revenue plan to zero and mean not delivering on the company’s strategy of continuous innovation in taste. Over the years, this strategy had solidified the company as a market leader, with overall sales in potato chips amounting to more than 10 billion dollars.

Luckily, the procurement innovation team had recently licensed Scoutbee. In their approach to identifying new suppliers, this meant having one tool instead of a plethora of sources to scan. Previously, each member of the team had been manually going through search engine results, supplier directories, Excel sheets with exports from supplier relationship tools, and even handwritten notes, but now they could sit together and discuss what they were looking for, based on the Scoutbee AI Supplier Search. 

This made it easier for them to translate the complex requirements formulated by R&D into precise search terms. The Scoutbee platform is based on feedback of thousands of supplier discovery cases. It features all the relevant angles a buyer needs to cover when expressing complex technical requirements as keywords, for example the product name, product type, application area, or the certificates a supplier is required to have. Once they had decided on these and other relevant keywords, the team could hit the search button and let the algorithm do the work for them. So, instead of investing further time into scrolling various sources, they could now focus on more strategic tasks while waiting for the results.

Within moments of entering their search, the team received a list of 24 suppliers that were totally new. Scoutbee’s AI technology had aggregated all the supplier data available on the internet relevant to the search keywords, analyzed it for suppliers with the highest match rate to the more than 20 search terms, and enriched this supplier data with information on revenue, employees, and certificates. Unlike traditional search engines that present a list of results based on paid advertisements, Scoutbee gives impartial results based exclusively on the user’s search criteria. In addition, crawling all available data at the start of the search ensured that the most up-to-date data was considered, circumventing the limitations of conversational AI applications like ChatGPT that run on a restricted data set. In further contrast to generative AI-driven applications, Scoutbee’s underlying algorithm is exclusively trained on procurement data, making the results more accurate and precise.
 

With the search results on Scoutbee, the team could now access the full profiles of each of the newly discovered suppliers, including location, number of employees, and certifications. Having this comprehensive view made having to go through various Excel sheets of supplier names (which, of course, included duplicates) and missing or outdated data a distant memory. As one of the buyers put it: “The speed and the accuracy of the search process impressed me the most.”

 

 

Moreover, the list and associated profiles meant the team only needed one meeting to discuss who to invite to an RFI. This process had previously taken at least four or five meetings due to uncertainty about having an actual view of supply options, as well as out of date or missing data, forcing the team to perform searches several times. Having a comprehensive view of the supplier landscape before the RFI offers another huge benefit: it avoids the risk of realizing you’re working with the wrong supplier further along the sourcing process. 

In summary, within only a matter of seconds, the procurement team was able to review a list of 24 previously unknown suppliers. Relying on the enriched profiles, they were able to select and invite a total of 17 suppliers for an RFI. Having completed the task ten times faster than previous manually-managed sourcing projects, and with higher quality results, the team felt confident in its ability to further deliver on the company’s strategy of continuous innovation.

Ultimately, equipped with these results, the procurement innovation team saved the day, receiving high praise from their colleagues in R&D - making it possible to bring the chips to the next round of product testing, where a select few product testers would get the pleasure of that first crunch.