• Written by Muhammed Abdulla NC, CEO

In theory, AI cost in procurement should be a straightforward calculation, but in practice, it is often a complex and multifaceted issue, with spend accumulating in unexpected places, and the true costs only becoming apparent after deployment. The business consequence of this can be significant, leading to budget overruns and reduced ROI. The mechanism behind this is the way in which AI systems are integrated into existing procurement processes, and the data work required to support them.

Integration and data work as the dominant line

The largest cost component of AI in procurement is often the integration and data work required to get the system up and running. This can include data cleaning, data mapping, and data migration, as well as the development of custom interfaces and APIs. For example, a procurement manager at a large manufacturing company may need to work with the IT department to integrate the AI system with the existing ERP system, and to ensure that the data is accurate and consistent. This can be a time-consuming and labor-intensive process, requiring significant resources and budget.

A shift supervisor at a distribution centre may also need to check the maintenance log to ensure that the AI system is properly configured and functioning correctly. This requires a deep understanding of the underlying data and systems, as well as the ability to troubleshoot and resolve issues as they arise. Most people get wrong about this is that they assume the AI system will be a simple plug-and-play solution, but in reality, it requires significant upfront investment in integration and data work.

Run cost against query volume

Another key factor in determining the cost of AI in procurement is the volume of queries that the system will need to handle. As the number of queries increases, so too does the cost of running the system, and this can quickly add up. For instance, a category manager comparing two suppliers may need to run multiple queries to determine which one is the best fit, and each query may incur a cost. The cost of running the system can be significant, and it is essential to carefully consider the query volume and the resulting costs when planning an AI deployment.

The cost of running the system can also vary depending on the complexity of the queries and the amount of data being processed. A simple query may incur a low cost, while a more complex query may incur a higher cost. It is crucial to understand the cost structure of the AI system and to plan accordingly to avoid unexpected costs.

The cost of a wrong answer nobody caught

A wrong answer that nobody catches can have significant consequences, including financial losses, reputational damage, and legal liabilities. The cost of a wrong answer can be difficult to quantify, but it is essential to consider it when evaluating the cost of AI in procurement. For example, a procurement team may use an AI system to determine the best price for a particular item, but if the system provides an incorrect answer, the team may end up overpaying for the item, resulting in financial losses.

It is also important to consider the potential consequences of a wrong answer in terms of reputational damage and legal liabilities. If a company relies on an AI system to make procurement decisions and the system provides incorrect answers, the company may be held liable for any resulting damages or losses. This can have significant consequences for the company's reputation and bottom line.

Attributing spend to a process rather than a model

When evaluating the cost of AI in procurement, it is essential to attribute spend to the underlying process rather than the AI model itself. This means considering the costs associated with data preparation, integration, and maintenance, as well as the costs associated with training and supporting the procurement team. For instance, a company may need to invest in data cleansing and data mapping to ensure that the AI system has accurate and consistent data to work with.

This approach can help to provide a more accurate picture of the costs associated with AI in procurement and can help to identify areas where costs can be optimized. By attributing spend to the underlying process, companies can better understand the true costs of AI in procurement and make more informed decisions about their procurement strategies. To learn more about optimizing procurement processes, visit our operations page.

When AI is the wrong choice

While AI can be a powerful tool in procurement, there are certain situations where it may not be the best choice. For example, in situations where the data is limited or of poor quality, AI may not be able to provide accurate answers, and the costs associated with data preparation and integration may outweigh the benefits. In these situations, it may be better to rely on traditional procurement methods or to invest in other technologies that can provide more accurate and reliable results.

If you are considering implementing an AI system in your procurement process and want to discuss the potential costs and benefits, please contact us to learn more. We can help you to evaluate the potential costs and benefits of AI in procurement and to determine whether it is the right choice for your organization.