“How much can I ask for it?” – that is the question. Reasonable buyer would not offer the highest possible bid in auction. Reasonable seller of professional services will not offer the lowest possible bid either. The author discusses client – provider engagement and pricing policy from the point of view of game theory. The article is intended both for clients, who are interested in getting the lowest price possible retaining the best service providers, and for vendors, who are motivated to get fair price for their services and withhold the price pressure.
Since the age of times to nowadays, pricing strategy remains to be a palpitating issue for market players on both sides of transaction. On complex modern markets (such as professional services market) the drive to sell high and buy low explodes into the whole universe of sophisticated questions. Can I win the contract by offering better price? What is my discount floor for this particular project or client? How should I carry out a tender between suppliers in attempt to find the lowest and reliable and high-quality offer? Nobody has heard yet about the one and the only correct answer. Understandably so – John Nash, who achieved impressive results in theory of non-cooperative games, has been awarded a Nobel Prize in 1994 not for nothing.
The idea to discover the rules of winning developed very far into sophisticated economical applications of game theory. The problems studied are very relevant and applicable to real world of professional services. Potential customers sometimes query prospective suppliers only once, and then other factors are taken into account. However, in most cases customer runs the first round of suppliers to narrow the search and select fewer suppliers to work with, then carrying out further negotiations with those initially pre-selected. Typically, prospective customer accepts offers in several rounds; on each one of them, he communicates the lowest bid to potential suppliers and asks them whether they want to come up with better offer. Such mechanism of finding the true value of the item is called an auction. Auction mechanism is also de-facto invoked to significant extent even in the absence of a single active seller, but with aggregate sellers of professional service market. In this situation, it is the market itself communicating “the current lowest bid”.
John Nash has shown (1950 and 1953) that suppliers competing for one and the same item in theory can try to negotiate with each other the limit of acceptable price. He also highlighted a very important detail: an opportunity to make a last minute better bid is always more attractive alternative for independent market player to loosing the contract. This drives the price further and further down to the level that constitutes equilibrium, which is below a sustainability level. This is due to amazing common consequence, proved to be true for all auctions:
Presuming that market players on average estimate the value of the item and their bid correctly, the winning bid produces lower than feasible or even negative profit.
Specifically, bidders tend to base their bids on unconditional expected value of the item. These bids represent their own estimates of value. We do reasonably assume that all bidders are professional service providers and therefore their estimates on average are correct. However, you only win if your estimate happens to be the highest of those competing for the item. Therefore, winning against the number of rivals following similar bidding strategies implies that the winner's estimate is an overestimate of the item’s value (or underestimate of feasible contract bid) conditional on the event of winning. This mind-boggling mathematical consequence is called the “winner’s curse” (you win, you loose, and you curse).
The winner’s curse results in less than acceptable or even negative profits for winners in all auctions. Winner’s curse is an intrinsic property of auctions of all types. It exists regardless of whether the bidders realize this danger or not. It is a consequence of the problem definition itself, rather than the bidders’ approach, tactics or behavior. At first glance, this conclusion invokes skepticism: it reveals counter-intuitive fact that bidders repeatedly commit mistakes in their own field. However, laboratory experiments show that even experienced executives immediately fall prey to the winners curse in auctions in other fields; numerous practical data show that the winner’s curse is a very pervasive phenomenon in real world, reported in many industries, from oil and real estate contracts to publication and distribution rights, and of course professional service contracts too. The winners curse is inherent consequence of adverse selection on the side of whoever owns the item to be sold, being the result of the seller’s sole decision rule. It is therefore completely out of buyer’s control. Winner’s curse is like hurricane or other act of Nature: attempt to change it is useless; you can only learn how to live when the circumstances look like it is coming.
Exactly how significant is the penalty of the winner’s curse for those unaware of it? Studies show that overly aggressive bidding translates into negative average profits for inexperienced bidders and losses for winners in more than 50% auctions. Experienced bidders less likely to see losses, but laboratory experiments still show that 41.6% bidders suffer from winners curse.i
So what is the strategy to follow for the supplier to work above sustainability level on one side, and for the customer to get reliable and supplier on another side? The answer to this question is undoubtedly important for all parties. Nobody is interested in driving the price below sustainability level, because in this scenario supplier will be deprived of capability to invest in business development, technology, quality and reliability improvement and will cease to exist. Customers, on the other side, will face considerable extra costs and risks which we describe later.
On one of the Dublin Localization summits carried out by Localization Institute, one very large, reputed and innovative client shared amazing story of how they tried to implement auction model in attempt to lower localization costs. Client engineers have devised a Web application where client managers could place localization contract lots and potential suppliers could place bids for these lots. Fortunately, localization managers who were the “owners” of the projects were left to make the final decision on whether actual sale of the lot is made, and closely monitored the auction. Both regular approved suppliers and new suppliers were allowed to participate. During the first auction round bidders indeed went quickly down below the usual suppliers’ price level, and what happened is that regular suppliers soon left the auction leaving only marginal and unknown providers to participate. Client managers have discovered that the last round leaves out only totally obscure and unknown to project owners companies to compete with each other. Project owners were protesting against placing the order with them, and managers decided to suspend trading and cancel the lot. (Notably, this decision was condemned as unfair and inappropriate by all auction participants, including even those who quit the game earlier.) To cure this problem, the rules have been changed so that the auction would continue only if at least one of regular approved suppliers still participates. In this new setup, however, regular supplier naturally won the auction because he has offered the same price, but had an advantage of much higher credibility with the client. Former outsiders were “used” in this setting only to draw regular supplier lower than usual. This clearly defeated the purpose of the auction, which was strikingly demonstrated when the lot was “sold” to one of the regular suppliers at lower-than-usual price. When the auction was completed, the winner has called the client and complained that the “excitement of gambling” has unwillingly drawn him below feasibility level and he is not actually able to take the job on board on these terms, “can we please renegotiate”.
The problems revealed by this case study fit nicely into the picture from other fields and our own experience. The scope of participants is the first most sensitive issue. If you open auction for everybody, then how you would ensure that the bidders qualify and actually are capable of meeting the quality and process requirements? If you pre-select the bidding audience with limited scope of approved providers, then again you are running into the same problem because all service providers are actually different and their price floor reflects their scale, costs and capabilities and you are not comparing apples to apples. The process repeats itself – whoever offers lower cost most probably offers different service. Even if we totally disregard individual differences and assume that whoever wins among approved providers is capable to deliver, the most probable outcome would be approved providers taking bids to the lower level and then supplying lower quality at this lower price.
There are also a number of other complications, such as for example the global suppliers are usually located in many different time zones, which makes it difficult and uncomfortable for some of them to participate in the auction interactively. Yet even more than that, as auction implies instant real-time decision-making process in dangerous break-even zone for the company, therefore only high level executives can possibly be authorized to represent their companies as qualified participants. This fact brings significant overhead on any company in this bidding contract placement model.
There are several important lessons to be learned both from the game theory and aforementioned case study, as well as from other fields. The first and foremost is that even as auctions are being very attractive as an idea, they are very difficult to implement as a working model, especially more so in the field of professional services. There are issues of scope of suppliers allowed to participate, there are issues of complex, changing and hard to formalize multi-factor selection criteria, and there is also a huge issue of unwanted cooperative behavior as well.
Undesirable cooperative behavior has been observed even in commodity auctions and may appear whenever a community of participants is formed. It is in the nature of humans to cooperate and make alliances, and it is in the nature of humans to use all means to achieve their goals. Unexpected cooperative behavior has been observed on large commodity auction marketplaces such as eBay. On eBay Motors, for example, car dealers initially used open auctions where buyers could chat and exchange information on the item and seller with each other. What happened, however, is that buyers have started to develop unfair cooperative behavior to exert extra money from dealers on extras, and even threaten and actually carry out defamation attacks to (even) blackmail and intimidate sellers. Now experienced eBay sellers generally use only closed auctions on high price tag items, protecting themselves from buyers learning and cooperative behavior which is often employed to bend the rules of the game in their favor.
Does this mean that the auctions should not be an instrument in professional service industries? Yes and no. Competitive sealed-bid auctions are commonly used in the commercial construction industry in USAii. However, considerable amount of private sector work, particularly for larger, more unusual jobs, is awarded through “negotiated” contracts rather than through auctions. This proves that with complex services and large projects buyers do appreciate the complexity of the problem and carry out a multi-factor analysis taking into account other considerations and complex indirect costs and values.
As mentioned before, the first key difference between services and commodity trading is that service transaction is not completed until the project is over and services are delivered and accepted, while commodity auction deals are done as soon as payment transaction and physical transfer of the purchase is completed. Extensive project duration basically means that the deal (and therefore renegotiation opportunity) remains de-facto open for quite some time. Another important difference is that there is an on-going relationship between client and service provider. These two differences provide at least three mechanisms of escaping the winner’s curse, either by claiming that “arithmetic errors” have been made, by asking for “help” from the customer, or else by submitting “change orders”. All three mechanisms allow the winner to escape the winner’s curse when bidding too low and immediately regretting their bid. These mechanisms are well known in construction industry and well described and developed, so the outcome in the mentioned case study is not an evidence of model deficiency, but rather a natural correction mechanism.
Relationship existing between buyer and supplier in professional services industry is not a marketing pitch – it truly represents quite tangible economic value. This value, specific to particular client, may vary and is difficult to estimate and quantify, however it can be evaluated. One can distinguish three channels where dollar value emerges:
- Customer wants his professional service supplier to be experienced and have a track record with his industry or company, because it saves time and money on training and endgame quality improvement.
- Customer wants his supplier to be reliable, to deliver on time and with good quality, so as to minimize the risk of rework, non-delivery and delays.
- It is also very comfortable for a client in professional services to trust his supplier, and uncomfortable to throw and wait whether the project will succeed or fail.
It would be rather shortsighted for the customer to believe that winner’s curse is a mathematical incarnation of the fact that interests of buyer and seller of professional services are not the same and buyer is not interested at all in any excessive indirect investment in supplier's business. Such a statement from customer side would mean that the customer does not understand the economical value of relationship between supplier and contractor, and therefore is not fully realizing all the risks and costs which are invoked when selecting the supplier on price only.
The conclusions we are drawing from all mentioned above concern both buyers and sellers. The customers should take into account the following:
- Correct implementation of auction contract placement mechanism requires significant research, extensive field experimenting and numerous safety checks and adjustment mechanisms. If developed correctly, this is going to be a very expensive toy which requires careful fine-tuning. If taken lightly, auction idea will immediately discredit itself due to the number of industry-specific and human factors which will invalidate the results. In the absence of fully developed and qualified engine, it’s best to stop the auction at some point where providers are short listed and place the contract manually, weighting all the factors.
- If you do not have yet well developed software or even process to take in and evaluate a competitive bid, take into account the following “rule of thumb”: the lowest price offer is definitely not the best one in service industries. Globalization went far enough to destroy tenfold price difference in different regions, and now professional services are roughly of the same hourly price in Bombay and Hong-Kong, Moscow and Jakarta. You can buy multilingual DTP services at anything between $3 and $10 per page for example, but the lower you accept below $3 per page, the more chances are that you are running into risks that you do not realize. Lower price often means that supplier overestimated his capabilities, is probably working below sustainability level and therefore has no reliability reserve or guarantee. This means that the quality will be lower than you expect and the work will take longer than planned. There is a risk that the work will not be delivered at all. It is not in the nature of average professional service supplier to openly and honestly admit that he is not capable of fulfilling the contract with due quality in this timeframe and within this particular budget. In most cases, he will simply deliver at his best leaving you to deal with this “deliverable”. It is then your own fault only that you have underestimated true cost of the required quality and actually ordered something less than you have been expecting.
- In any case, the price should not be the only criteria for selecting the supplier. You must remember that different suppliers actually provide very different services. They differ in all aspects, from communication (or lack thereof) to experience, knowledge and scalability, to quality of deliverables and many other aspects that are very difficult to quantify. This simply means that more expensive, overqualified supplier can prove to be actually more profitable to work with because he is more experienced, or better reacts to unexpected situations, or offers some unexpected benefit that you may suddenly need. There are extra costs involved in working with cheaper suppliers – the cost of additional monitoring, the risk of rework cost and cost of product release delays.
- By selecting the offer above the lowest you are not overpaying, you actually increase the reliability of project completion, decrease your indirect costs and manage the risks.
- You have to know your supplier well. Invest into research of his actual capabilities. If you like the supplier but think that his price is a little bit high, work with him further to negotiate the price to the level that is acceptable for you.
The suppliers, on their side, have to remember the following:
Generally, the answer to overcome the winner’s curse is to determine the value of the prospective contract conditional on the event of winning, rather than try to determine its value unconditionally. This means that experienced supplier avoids winner’s curse watching vigilantly for the bid price being above sustainability level as internally defined to be acceptable for his business. This slightly decreases the probability of winning, but in return offers reliable protection against financial losses due to the ill-set prices. Experienced supplier knows that he is not willing to win the contract if it means working below sustainability level. Experienced bidders lean a set of situation-specific rules of thumb that help avoid overbidding in the real field setup. (What is interesting that these rules are very situation and industry-specific, to the extent that experienced bidders in one field fall prey to winner's curse in auctions in other fields; this is well demonstrated in laboratory experiments.) By all means, try to avoid “bare”, commodity-like auction, where the price is the only criteria. Company with sensible pricing policy has little chances of winning on price only; and trying to win by price dumping is deficient and destroying practice. You may well find yourself going further and further down only to discover suddenly that you have descended below sustainability level before you knew how you did it. I would even say that the client who regards that the price is the only criteria is not worth to work with and fight for.
- It is extremely important to acknowledge your costs correctly, and under no circumstances accept contracts that are below profitability level, because:
Whenever possible, try to explain to the prospective customer how your services are different from the others. Truly strong and lasting relationship between customer and provider and stable equilibrium market is based on those unique features that make providers different from each other. It is the variety of services that helps to create stable and mutually profitable equilibrium of the parties pursuing different interests of their own. It is of critical importance to position yourself on the market correctly, actively demonstrating to customers and competitors the unique features that you have and that actually make it legitimate for you to charge what you charge and continue to be engaged in your field in the situation of fierce competition.
- Whatever customer says, there is no guarantee whatsoever that the contract will be extended, and therefore the attempt to “buy in” client relationship by offering the price below sustainability is an extremely risky investment;
- It is next to impossible to increase rates and return to acceptable level. Even if you manage to preserve the relationship, all chances are that you will have to work with him on the same lower prices.
It goes without saying that we are not possibly able to cover all possible setups in all their details, as well as their models and numerous details and conclusions. We only attempted to come up with very general principles applicable to non-cooperative setups on professional services markets. I do believe that this field will develop further and new software will be created soon to demystify and automate the multi-factor decision-making process of proper contract placing for complex professional services, taking into account various specific details. It is time to get yourself prepared, regardless on what side you may find yourself.
i Garvin, S., and John H. Kagel. (1994). “Learning in Common Value Auctions: Some Initial observations.” Journal of Economic
Behavior and Organization 25:673-82.
ii Dyer, Douglas and John H. Kagel. (1996). “Bidding in Common Value Auctions: How the Commercial Construction Industry Corrects for
the Winner’s Curse”, Management Science. 42:1463–75.