Twenty years ago, we were forming GPOs and Purchasing Consortiums to take advantage of Economies of Information. Ten years ago, we were not only rolling out suites that focussed heavily on consolidating spend (and performance) data globally (with the likes of Coupa and Sievo boasting about how much spend they had managed and normalized) to take advantage of the economy of information global spend data gave us, but building best in class analytics solutions to take advantage of all of the data they (could) gather(ed).

In the age of (predictive) analytics, which preceded the current age of AI Hype, the importance of data finally started to become recognized and you had a number of startups hit the scene providing next-gen data feeds. (Near) Real-time commodity indices, market price data, risk data, company financial data, carbon data, energy rates, water rates, regional overheads, average process time, average GPO and transaction rates, average performance data, etc. Any organization that wanted to build a best-in-class should cost model, best in class performance model, etc. The information was available, there was an economy for information, and the economy of information was right around the corner.

Let’s step back and define what we meant by this. On economies of information, twenty years ago we wrote:

The consortium of the future offers the benefit of expertise more so than it offers the benefit of scale. Eventually, especially with constantly rising raw material prices, the best practices employed by a competent consortium will squeeze all of the fat out of the supplier’s margins and the best price will be obtained. Once this occurs, the consortium will use its expertise to assist its members in advancing purchasing technology, reducing wasteful consumption, and improving the application of the goods and services they purchase. Since a consortium has access to all of the knowledge of its members, it can tap this knowledge to identify the best potential suppliers with the best potential products and services to meet member needs. Furthermore, this gives it a much better chance of identifying and qualifying low risk suppliers.

In other words, with a fact-based outlook on reality, consortiums could help take Procurement to the next level. Then, when the data-stream startups made all of that same information easily available as plug and play data feeds into your suite through standard APIs, the true economy of information hit Procurement for those who wanted it and Procurement could make insight-based and fact-based decisions and get better.

But now that we’re a few years into the age of AI Hype, I believe we’ve lost the economy of information. There a few reasons for this:

• we’ve replaced data feeds with LLM chatbots like clod and chat, j’ai pété and assume they have access to the same data, and, most importantly, the same ability to run predictive analytics on that data
• despite claims to the contrary, the LLMs are getting worse by the day … now that the majority of data on the internet is AI generated slop, being cross fed into other LLMs, regurgitated with compounding errors, we are not only losing the core data in the tsunami of slop but the meaning of that data as well
• with LLMs being cheaper than data feeds, the data feeds have been ignored, a number have went out of business, and the rest are floundering

There’s no information without actual, verified, facts and intelligent interpretation, and the majority of that has been lost in the age of AI hype.

If too many real data providers, as well as applications that deterministically and intelligently integrate and analyze real data and real facts, go out of business, there will be no solid foundations for real information, and, thus, no solid foundations for economies of information — and then we’ll be back to the Procurement dark ages.

Technology has never advanced Procurement. Only facts, data, process and decision improvement based on intelligent interpretation has.