Paddy Englishman, Paddy Dutchman and Paddy Japaneseman Walk into a Supplier Relations Bar

We are pleased to bring you a post on our Supplier Relationship Management hot topic this month from Dr Alan Holland, Founder and CEO, Keelvar - sourcing optimisation and advanced eAuctions specialist.

Growing up in school in Ireland, Paddy Englishman, Paddy Scotsman and Paddy Irishman jokes were always popular. Invariably, it was Paddy Englishman who was always the butt of the jokes but then a boy from England moved to our rural Irish school and told a joke in which Paddy Irishman suffered the punchline. There was stunned silence. We informed him that this was not the way the joke worked.

He then made a good point, nobody in England was called Paddy so Paddy jokes worldwide were all aimed at Irish people. There was then an extended silence but someone interjected “what about Paddy Ashdown, he’s the biggest joke of all.” We all laughed and agreed to disagree, ignorance was such bliss.

So the title of this article refers to Paddy Englishman, Paddy Dutchman and Paddy Japaneseman which is in reference to the three kinds of eAuction format most commonly used in procurement. So which variety of reverse auction is the butt of the joke in eSourcing and supplier relations? The following table offers a short reminder of the key design aspects of each reverse auction before we analyse it in a little more depth.

Reverse Auction Description
English The buyer may set a reserve price above which she will not buy the good/service and sellers make bids. It is open-outcry since everyone is aware of every bid made, and it is first price since the buyer pays the lowest winning price. The bids are made in descending order.The auctioneer is calling the bids, the end of the auction is when no bidder accepts the suggested price. It is also possible to fix a time limit for the receipt of final bids, either as an absolute deadline or relative to the time at which the last bid was received (e.g. 5 minutes after the last bid is received).
Dutch The difference between Dutch and English reverse auctions is that the bids are ascending rather than descending. In a typical Dutch reverse auction, the auctioneer starts at a low price below which anyone is willing to pay, and then increases the price. The first bidder to accept the price gets the good.  If the price increases below the reserve price, there is no sale. The Dutch reverse auction is also known as the “ascending clock” auction since, in practice, the auctioneer indicates the price using a large clock-like dial with the price ticking upwards.
Japanese By combining aspects of the English and Dutch auctions we get Japanese auctions. Here the auctioneer calls out descending prices and bidders drop out of contention when the price gets too low for them. The last bidder remaining wins at the price.


The workhorse of the auction arena is the English auction as it clears efficiently, doesn’t impose onerous information disclosure on bidders and has a proven record across all industries and spend categories. It’s clearly not the butt of any joke so is there anything wrong with the Dutch or Japanese auctions that could give us a clue as to why they are less popular? The following table examines the most appropriate occasions for using the respective auction varieties and so begins to uncover key factors.

Reverse Auction Appropriate Use Cases
English This is the most general auction clearing mechanism that should be used in most sourcing exercises that are price focused and don’t have a tight closing deadlines.
Dutch When timing is critically important, e.g. the procurement of perishable goods such as flowers, meat, fish etc, then the Dutch auction is most effective. It leads to wider variance in clearing costs but on average produces the same cost as the English auction. But the English auction doesn’t usually take too long to clear either so the Dutch auction is only marginally favourable when time is a major issue.
Japanese Never use them! Why? Because this form of auction expects bidders to reveal their lowest possible cost for delivery of the good or service even when they are not in the top two bidders. This feels like an unfair and unwarranted intrusion so suppliers take a negative view of buyers who try to apply this format.


So the butt of the joke is the Japanese auction because it should never be used in practice by responsible buyers who wish to maximise bidder participation or maintain good relations with suppliers. Suppliers attach a high cost to information disclosure when competing in Japanese auctions so it is likely to reduce competition and lead to more inefficient outcomes.

So why do some auction technology vendors offer Japanese auctions at all? This is a good question as the arguments in favour of Japanese auctions are vague, unscientific and exceptionally weak.

Share on Procurious

Voices (4)

  1. REY:

    Hello all,
    I agree with both of you.But I would add a very simple rule of thumb:
    – english auction (with graphic price visibility if possible or multi-criteria) when there is a high number of bidders. You need to show bidders there are facing many competitors. Ranking is unefficient and I would rather avoid.
    – Dutch (or Japanese) auction when you have a small number of bidders. The bidders do not “see” each other in these formats. They don’t even know if there are competitors on the other side. Although I would not recommend it, these 2 formats may work with only ONE bidder !

  2. Nick @ Market Dojo:

    Hi Alan,

    Thanks for the reply. I agree that there is a tendency for savings to be over-egged, though it’s a side-issue to our commentry 🙂

    Apologies, I wasn’t very clear on the BAFO side. The client conducted that exercise prior to any discussion of an auction, or indeed any further negotiation. It literally was a request for a final offer. There was no expectation of later opportunties, hence why I classed this as a BAFO.

    Whilst we’d love to repeat the exercise and follow a Ranked auction format, as I mentioned, this is a scenario that would not work for a Ranked auction. There were too few bidders per individual price element plus it was to establish neogtiated rates for a framework. There are no winners or losers, so ranking is meaningless. Much like the rebate example I mentioned too.

    One final note on sullying supplier relations: after every auction run in our tool, we automatically ask bidders for their feedback via a simple question ‘Between 1 and 10, how much would you recommend us?”. We had 13 survey responses after the Japanese auction, 5 of which were a score of 10 out of 10 with comments like “Very simple process”, “very good” and “easy to use”.

    Based on our objective data, as opposed to subjective views, we would say that the feedback from suppliers on Japanese auctions is on a par to that of Ranked auctions. At least, in our software anyway.

  3. Alan Holland:

    Thanks for the comment.

    In general, I think the procurement profession is in danger of overstating claims on savings because they are intrinsically impossible to measure with precision. For example, when one states that the outcome of Stage 1 in a sourcing exercise was $x and the outcome in Stage 2 was $y and because x>y this measures ‘quantifiable savings’.

    In your example you stated that ‘Best and Final Offer’ was superseded by another superior offer. Then it wasn’t a final offer at all… The integrity of the process was ruined by the knowledge or expectation of latter rounds. Of course people enter stage 1 with high bids – that’s the rational thing to do.

    The key thing to judge the efficacy or otherwise of a Japanese Auction would be to rerun the process with an English Auction. Even if bidders didn’t attach a cost to information disclosure (which they almost certainly do) then the equilibrium for English Auctions is to have a narrow distributions of clearing prices. And it doesn’t sully supplier relations. Wedded to the fact Japanese Auctions will have lower participation because of the informational costs, then you would have been better off running a straightforward English Auction (there’s no need for the BAFO stage – that adds no value).

    I concur that auctions have a lot of value to offer – there’s plenty of scientific evidence to back that up. But we’ll have to differ on the efficacy of Japanese Auctions.

  4. Nick @ Market Dojo:

    Great timing on the article! Having just witnessed a client save in excess of £1m through a Japanese auction, I would say that, like any eAuction, there is a time and a place for it in the arsenal.

    Firstly, why use a reverse auction at all? You could just ask for BAFOs and be done with it. That is what our client did prior to coming to us. However, they elected to do the Japanese auction on top, and it delivered an additional £300k saving. Across 120 bidders and several hundred price points, it would have taken a long time to do this via email/phone/face-to-face.

    So, could a Ranked Auction have been used? Well, Ranked auctions are generally ineffective when you have low liquidity, i.e. one or two bidders per price element. Japanese auctions overcome this challenge.

    Ranked auctions make little sense when you are not auctioning committed business, but are negotiating rates to feed into a framework agreement. It doesn’t matter if you’re in 1st, 3rd or 10th and that would just obfuscate the process. However a Japanese auction serves a purpose here, as it drives better than BAFO (as per the real example above) to feed into an agreed schedule of rates for future call-off, without any confusion around market positioning.

    Further to this, Japanese auctions also are fantastic for negotiating rebates across hundreds of suppliers, many of which aren’t even competitors, as they could be from many different supply categories, so you couldn’t possibly ‘rank’ them against one another. Instead these hundreds of suppliers have the opportunity to put forward their negotiated rebate agreement in isolation, yet driven by the bid momentum created by the Japanese auction.

    Ranked auctions can often be undermined by ‘bid shadowing’, where an incumbent can closely follow the leading bidder, knowing that if they stay sufficiently close to them, yet also not giving away as much margin, it is likely they will retain the business due to the hassle faced by the buyer of switching. Japanese auctions overcome this challenge too. Plus it gives buyers the benefit that suppliers are not aware of their final position, hence buyers won’t face a disgruntled supplier who knows they won the auction but is not being awarded the business.

    Just to clarify we are relating to the private sector here and not the public sector, who obviously have defined rules about what suppliers must see during an auction.

    So I hope that I have given some examples of why a Japanese auction creates additional opportunites for buyers. Afterall, the article in question could equally apply to auctions as a whole, as there are certainly some in the procurement industry that feel that the argument for a reverse auction full-stop is “vague, unscientific and exceptionally weak.” However, many other respected professionals see through this and would beg to differ.

Discuss this:

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.