THE PROPHET, who has recently discovered ranting is his new favourite thing to do (on LinkedIn), recently complained that Procurement, Commodity, and Supplier Data is Too Darn Expensive.

And while he’s right in that data is often too expensive for what it is, it’s not going to stay that way. Next generation providers are going to commoditize quality data and lower anonymized community data subscriptions to win (and keep) clients, because they know that there’s no value in advanced technology alone (and especially in analytics, optimization, and AI wihtout quality data to feed it) but there are three key points he missed in his rant where he complained about data prices and advocated the use of LLMs and Gen-AI as a substitute (which they are not, and considering how much they hallucinate, we wouldn’t even trust them to be directionally accurate — just feed the historical data you can get your hands on into Excel and do some basic trend plotting if directionality is enough).

1) As Lisa Reisman noted in the comments, sometimes you need highly granular accurate data by geography, volume, and production methodology. When pennies make a difference, because you are buying tens or hundreds of millions worth of the material for a global operation, it matters.

2) Most firms are still ignoring their own data, which, when run through something like Covalyze (which THE PROPHET should love as it was founded and designed by economists), gives very accurate target cost models on any category the firm has enough historical data on, allowing them to pinpoint where they need more data and why for cost breakdowns (and should cost models to refine the target cost models), and which suppliers they actually need those expensive profiles on. Then they can go to pay by the sip providers like Veridion for basic supplier data or other emerging commodity and supplier data portals.

3) The amount of data most firms need is much less than they think. In the tail, most of the spend is not significant enough for any market data to provide insight on a significant savings potential beyond what you will get from analyzing your own historical data and market quotes. When pennies won’t make a difference, you don’t do detailed cost breakdowns by raw material. When the product is a commodity that can be supplied by multiple suppliers at similar price points and equal quality levels, you don’t do deep risk profiles because you can just go to the next supplier in the queue if the first one fails you. And so on. You only do detailed analysis where there is statistical likelihood of a real opportunity or a real risk. Otherwise it’s a waste of time, money, and resources as no organization today even comes close to fully analyzing the significant categories and risks they have in any given year. Thinking you will do more is delusional and not worth it if you don’t have the basics covered.

By the time firms actually need more data, you can bet a next generation of data providers will have it readily available and cheap by today’s standards.