Blog Post
November 13, 2019

When algorithms support buyers in everyday tenders

Written by
Tim Grunow

When Algorithms Support Buyers in Everyday Tenders

Imagine you are running a trucking business and you need to load a truck with boxes of different sizes and shapes. Moreover, the boxes need to be unloaded at different locations and traffic needs to be taken into consideration, everything to ensure shortest possible delivery times. Things quickly get complex. In strategic sourcing, the boxes are different supplier offers and the optimal (un-)loading and routing represents the most desired awarding scenario where financial savings are maximized while also all internal and external non-price constraints are being considered. The process of getting to this most desired outcome is called optimization.

Whether you are working in logistics/freight, packaging, commodities or raw materials purchasing, sourcing optimization enables buyers to explore untapped savings opportunities and opens their minds to plenty of possibilities. However, optimization is still often under-utilized and procurement teams analyze supplier bids in Excel to find the best purchasing scenario - often combined with the buyer’s gut feelings. In the cases where optimization is used, it is not applied to all different levels of sourcing complexities, but rather for the most complex tenders only. One of the reasons for this lack of use are tedious setup times and missing guidance inside the tool.

These reasons hinder buyers from starting to adopt currently existing solutions in their everyday sourcing. That is why we at Archlet believe traditional sourcing optimization is dead. Buyers want to be guided by intelligent algorithms in the most intuitive manner; and optimization is just a means for that.

For the purpose of visualizing the mentioned insights on current adoption of optimization, we have drawn a pyramid (see below) representing all the tenders that a company launches over the course of a year and their corresponding complexity (on the y-axis). Some tenders are less complex and therefore run on a more regular base (weekly, monthly). Others are more complex and less frequent. The orange area at the tip of the pyramid marks the percentage of tenders where companies currently (would) put up with the long and tedious setup and configuration times, because spend is typically very high (> 30m). The orange area at the bottom of the pyramid marks low complexity, single-source tenders, where for example due to legislative restrictions no competitive business can be won. These processes are repetitive and can easily be automated. For all the other tendering processes between the two orange marked areas, optimization is still largely under-utilized — but offers great potential if it is guided and easy to use.

Pyramid showing where technology solutions can support sourcing processes based on the complexity of tenders

One of the mentioned reasons for low adoption of supporting software is due to the fact that existing solutions are process-centric rather than user-centric. Imagine if using intelligent sourcing software tools was as easy as using Instagram or WhatsApp. This in combination with the needed guidance will ultimately be one of the keys to higher adoption of algorithms, supporting buyers to increase efficiencies and savings across all spend categories.

Coming back to the tender pyramid, we see that by driving down the orange area through higher adoption of intelligent algorithms, we can tackle more spend of companies in times of ever-increasing pressure of expected year-on-year savings. And ultimately intelligent optimization is the driving core for smart automation in procurement. Other intelligent data modules allowing buyers to determine how and when to source will drive the next generation of cognitive procurement solutions.


Pioneering thoughts on data-driven sourcing