One question precedes every business case, and it is rarely asked. Why should the same demand, with the same suppliers, lead to a different price simply because the tender runs through a tool instead of spreadsheets and email?
The answer is not self-evident. A tool does not negotiate, it buys nothing and it does not know the market better than the buyer. If it nevertheless influences the price, it must do so through mechanisms that can be named.
Two of them are particularly relevant for a buying group.
A supplier does not price a bid at cost. They add a mark-up. Its size is the result of a trade-off. A high mark-up means more margin if the order comes, but lowers the probability of winning it.
Economic research on tendering has described this relationship for decades. Riley and Samuelson showed in 1981 that the mark-up falls as the number of competitors rises and approaches zero with many bidders.
What matters is not how many suppliers were actually invited, but how many the individual supplier assumes are bidding. The mark-up is a response to an expectation.
This is where the sharpest difference lies between an email enquiry and a structured tender. Whoever receives an enquiry as an attachment usually does not know whether other suppliers were asked at all, how many there are, and whether their bid will end up structurally comparable. They may hope that the existing business relationship, the buyer's convenience or simply the absence of comparability will protect them.
That hope is an amount in their calculation.
In a structured procedure both uncertainties disappear. The item list is identical for all suppliers, so bids will be comparable. The involvement of several suppliers is visible, so competition is real. The supplier has to assume the others are pricing tightly, and adjusts the mark-up accordingly.
The best-supported finding in the research concerns the number of suppliers that actually submit a bid.
Jeremy Bulow and Paul Klemperer showed in 1996 that a tender with one additional serious bidder yields the buyer more on average than an optimally conducted negotiation with the existing number of bidders. In their phrasing, one more bidder is worth more than the best negotiation strategy.
Applied to procurement, this means the effort of bringing one more suitable supplier into a procedure delivers more on average than any improvement in negotiating with those already involved.
The connection to the tool therefore lies not in the price but in the cost of that additional invitation. Whoever runs a tender through spreadsheets and email pays for every further supplier with additional work. Documents have to be sent once more, responses sorted once more, and the deviating structure of the additional bid aligned by hand once more. In a structured procedure the additional supplier is an entry in a list.
So a tool does not shift the price, it shifts the cost of creating competition. The price effect follows from that.
For a buying group there is an amplifier that a single company lacks.
Both mechanisms presuppose that suppliers are interested in the order at all. At 55,000 euros of annual volume in a category, that is not the case. Above one million euros it is.
Bundling creates the precondition, the procedure exploits it. Whoever has only one of the two leaves the larger part of the effect on the table.
What is expressly not meant here is collusion between suppliers. The effect described rests on each supplier pricing individually while factoring in the others. The payoff structure corresponds to a prisoner's dilemma. If all applied a high mark-up, all would be better off, yet each one improves their own position by undercutting the others. This structure is the reason competition in procurement works at all, and it exists independently of the tool in use. A tool does not create it, it merely removes the uncertainty about whether you are in it.
Does this hold without an auction? Yes. In the research, the ordinary tender with sealed bids is the standard case, not the special case. No auction needs to be involved.
Our suppliers have known us for years. Does a procedure change that? Not the relationship, but the pricing. The mark-up responds to expectations about competition, and those expectations change as soon as the involvement of several suppliers is visible.
What about categories with only two or three possible suppliers? There the mechanisms described apply only in a limited way. That is one of the constellations in which a tool does not pay off on the numbers, and we name it explicitly in the study.
Chapter 3 of the study describes four mechanisms with the underlying research and three limitations, two of which argue against using a tool.
→ To the white paper
The answer is not self-evident. A tool does not negotiate, it buys nothing and it does not know the market better than the buyer. If it nevertheless influences the price, it must do so through mechanisms that can be named.
Two of them are particularly relevant for a buying group.
The mark-up you do not see
A supplier does not price a bid at cost. They add a mark-up. Its size is the result of a trade-off. A high mark-up means more margin if the order comes, but lowers the probability of winning it.
Economic research on tendering has described this relationship for decades. Riley and Samuelson showed in 1981 that the mark-up falls as the number of competitors rises and approaches zero with many bidders.
What matters is not how many suppliers were actually invited, but how many the individual supplier assumes are bidding. The mark-up is a response to an expectation.
This is where the sharpest difference lies between an email enquiry and a structured tender. Whoever receives an enquiry as an attachment usually does not know whether other suppliers were asked at all, how many there are, and whether their bid will end up structurally comparable. They may hope that the existing business relationship, the buyer's convenience or simply the absence of comparability will protect them.
That hope is an amount in their calculation.
In a structured procedure both uncertainties disappear. The item list is identical for all suppliers, so bids will be comparable. The involvement of several suppliers is visible, so competition is real. The supplier has to assume the others are pricing tightly, and adjusts the mark-up accordingly.
Why an additional bidder beats a better negotiation
The best-supported finding in the research concerns the number of suppliers that actually submit a bid.
Jeremy Bulow and Paul Klemperer showed in 1996 that a tender with one additional serious bidder yields the buyer more on average than an optimally conducted negotiation with the existing number of bidders. In their phrasing, one more bidder is worth more than the best negotiation strategy.
Applied to procurement, this means the effort of bringing one more suitable supplier into a procedure delivers more on average than any improvement in negotiating with those already involved.
The connection to the tool therefore lies not in the price but in the cost of that additional invitation. Whoever runs a tender through spreadsheets and email pays for every further supplier with additional work. Documents have to be sent once more, responses sorted once more, and the deviating structure of the additional bid aligned by hand once more. In a structured procedure the additional supplier is an entry in a list.
So a tool does not shift the price, it shifts the cost of creating competition. The price effect follows from that.
Why both work harder in a bundle
For a buying group there is an amplifier that a single company lacks.
Both mechanisms presuppose that suppliers are interested in the order at all. At 55,000 euros of annual volume in a category, that is not the case. Above one million euros it is.
Bundling creates the precondition, the procedure exploits it. Whoever has only one of the two leaves the larger part of the effect on the table.
What is expressly not meant here is collusion between suppliers. The effect described rests on each supplier pricing individually while factoring in the others. The payoff structure corresponds to a prisoner's dilemma. If all applied a high mark-up, all would be better off, yet each one improves their own position by undercutting the others. This structure is the reason competition in procurement works at all, and it exists independently of the tool in use. A tool does not create it, it merely removes the uncertainty about whether you are in it.
Frequently asked questions
Does this hold without an auction? Yes. In the research, the ordinary tender with sealed bids is the standard case, not the special case. No auction needs to be involved.
Our suppliers have known us for years. Does a procedure change that? Not the relationship, but the pricing. The mark-up responds to expectations about competition, and those expectations change as soon as the involvement of several suppliers is visible.
What about categories with only two or three possible suppliers? There the mechanisms described apply only in a limited way. That is one of the constellations in which a tool does not pay off on the numbers, and we name it explicitly in the study.
The next step
Chapter 3 of the study describes four mechanisms with the underlying research and three limitations, two of which argue against using a tool.
→ To the white paper