Waiting For A Tsunami Of Text

In a world where text can be generated by an algorithm for negligible costs, we have to rethink how we value the written word. In procurement, the effort required to create written texts has become a proxy for validity. That is changing, in a world where Chat-GTP is the dunce in the room, ie. 2025, there will be little or no intrinsic merit to a requesting a response written across multiple pages because any algorithm will be able to do it. Adopting this approach will just result in more work for those assessing the bids.

One way to handle the incoming wave of computer generated text is to restrict who can bid, but you need to be careful not to restrict competition, as we need competition to foster innovation and deliver value. Instead, here’s four things buyers can do to adapt to an AI future, without resorting to anti-competitive practices:

1. Make bids easy to evaluate. The UK’s digital marketplace often uses the 100 word answer approach to evaluate multiple bids. By restricting the word count bids are easy to score and suppliers can be quickly whittled down do just three or four suppliers from only four or five questions. Taking dozens of suppliers down to handful using easy questions is better for suppliers and buyers.
2. Truly understand your needs. Too often buyers publish tenders instead of researching a market. This is like taking a wild swing and hoping to hit the ball out of the ground. AI can be used as a mechanism to research and evaluate markets,. Good data can be used to find innovative solutions and new competitors. Better research can help buyers narrow in on their precise needs and then going to market for those needs.
3. Monitor supplier performance. One bad result doesn’t necessarily make a supplier bad, but if we can make it easy to evaluate a supplier’s performance across dozens of contracts, well that gives buyers a much better sense of whether they can deliver.
4. Use more than text. Bidders can be evaluated in an interview, so why not evaluate them in a video too? Or maybe evaluate them in a day long interview with practical tasks working amongst your team? Or maybe ask them to produce a sample of code, or ask to visit one of their existing clients with them? There are plenty of reasons why we use text, but in a world where text is no longer the differentiator, we’re going to need to think about other ways to evaluate bidders.

The future of procurement doesn’t need to be so different, but it does need to change if we’re not going to be drowned in a wave of algorithmically generated text.

Spend Network have government procurement data for more than 150 countries globally. We take the hard work out of collecting, organising and analysing procurement data at scale. Get in touch to find our more about our data products and services.