To show how Inspire works, a simple negotiation has been set up with the objective of trying to secure a contract between two companies, Rosa Inc. and Casa Ltd. Rosa wants to sell an aircraft which Casa is considering purchasing.
Two Inspire users negotiate for Rosa and Casa. Like many Inspire users, they have decided to use pseudonyms rather than their real names. "Misty" negotiates on behalf of Rosa Inc. and "Smiley" represents Casa Ltd. After they each request an Inspire account and are paired up by the system, they are ready to begin.
Negotiations between Misty and Smiley
The parties and the problem
Both Misty and Smiley have carefully read the information about their respective organizations to understand the problem and its issues. Since each negotiator can access the system and make a proposal independently, we will follow Misty's side of the negotiation.
There are only two issues in this simple negotiation: the price of the aircraft and the terms of the warranty. It has been established that the normal price of this aircraft is in the range of $300 000 to $320 000. The sensible increase is of $10 000. Thus, the price options are $300 000, $310 000, and $320 000. In this industry there are four types of warranty typically available. The options are: no warranty, a 6 month, one year, and a 2 year warranty.
Both negotiators analyze the two issues and their associated options in terms of their relevance to their respective organizations and move to the pre-negotiation phase.
To prepare for the negotiations Misty and Smiley each rated the two issues. Note, that the pre-negotiation steps are conducted independently; one negotiator can never see the information (ratings) that the other negotiator enters.
Misty feels that price is far more important than warranty. Therefore, she assigns 70 points to price and 30 to warranty. Although Misty does not know it, Smiley feels that each issue is equally important and so Smiley assigns 50 points to each.
Each issue has one or more options, for example, price has three options: 300 000 $, 310 000 $, 320 000 $. After rating the issues, the options in each issue must also be rated similarly. In the Inspire system, for each issue at least one option must be assigned the maximum rating for the issue and at least one option must be assigned a rating of zero.
Misty considers the three options for the price of the aircraft and assigns the maximum rating (that is, 70) to the price of $320 000 because Misty represents Rosa which wants to sell the aircraft. The lowest possible price is assigned a rating of zero. Misty considers the price of $ 310 000 as somewhat acceptable and assigns a rating of 45.
Misty assigns ratings to each warranty option in a similar way. Note that for Misty, "no warranty" has the same maximum rating (30) as the 6 months warranty perhaps because the organization does not think it is possible that this plane will fail during the first 6 months of operation.
Given user's ratings for each issue and each option, Inspire calculates ratings for complete packages that are the subject of negotiations. A package consists of price option and warranty option, for example, "320 000 $ and No warranty" is one complete package. Inspire presents a few packages and their ratings so that the user can assess if the Inspire's results accurately describe the user's preferences.
Misty's evaluations of four packages are below:
The above ratings indicate how good the packages are given Misty's ratings of the issues and options. Comparing the packages Misty decided to change two ratings, for the second and fourth package. The second package while worse than the first (best) package is still quite good so Misty increases its rating from 80 to 82. The fourth package is not nearly as good as the third package and Misty downgrades its rating to 69.
Inspire uses now the issue, option and package ratings and it determines the utility function which closely reflects Misty's preferences. These function will be used to provide a rating for every considered package.
Exchange of offers and messages
Misty thought a while on how to begin negotiations. In the meantime Smiley prepared and sent the following offer together with a short message:
Misty considers Smiley's offer and -- contrary to Smiley's expectations -- finds it unacceptable. Note that the offer's rating is 0 (Misty's rating: 0). Misty prepares a counter-offer which better reflects Rose Inc. requirements. A message is attached to Misty's offer.
Note that the rating of the above offer is 95. From Misty's ratings of Price and Warranty (see section "Option rating") it would appear that Mysty's offer "320 000 $ and 6 months" should be rated 100. However, the adjustments that Mysty made in the Package evaluation step caused a drop in rating to 95. Although these adjustments were made only to two packages they affect ratings of all possible packages because they modify Misty's utility function.
At anytime during the negotiation, a graphical overview of the history of offers can be viewed. The graph plots the the users' ratings of offers sent (in green) and received (in red) and the time that the offers were sent. The graph can be viewed from the "View offer and message history" link.
Below are two graphs depicting the history of negotiations. The graph on the left-hand side represents the history from Misty's point of view because Misty's ratings is used. Note that Smiley cannot see this graph. The graph that represents Smiley's perspective is shown on the right-hand side.
We show the two graphs to indicate that:
- the two graphs are different because each represents a different point of view;
- sometimes what one side considers as a concession is not a concession from the other side's point of view.
Compare Smiley's two offers: Offer 2 and Offer 3. In Misty's graph Offer 3 is worse than Offer 2. This is because in terms of Misty's utility function Offer 3 has lower rating than Offer 2. Misty may think that Smiley does not want to negotiate because Smiley makes a negative or reverse concession (proposes an offer that is worse than proposed previously).
While Misty cannot see the graph on the right-hand side, Misty knows that such a situation is possible. Indeed, Smiley thinks that a small concession has been made on behalf on Casa Ltd. because the rating in terms of Smiley's utility function for Offer 3 is lower than for Offer 2. The graph on the right-hand side clearly indicates this.
Both of our negotiators should be well aware that such a situation may happen. What one side perceives as a concession, the other side may not see as such. Worse, the other side may consider this as a "reverse concession" or a hardening of one's position.
Misty and Smiley continue to exchange offers and messages. On one occasion Misty decided not to send an offer but ask Smiley for clarification. Thus, Misty sends only the following message:
Misty and Smiley exchange the offers and messages until, in the fourth round, Misty presented the following offer:
Smiley reviews Misty's last offer which is shown below. (Note that the offer's rating now reflects Smiley's preferences).
After short consideration Smiley accepts Misty's last offer. Smiley does it by selecting, on the Inspire page, the option:
— Yes, I accept my counterpart's most recent offer as listed above.
In some negotiations it may happen that the parties reach an agreement but there is one or more packages which are better than the accepted offer for both sides. Note, that better is measured with the parties utility functions. Thus, there may be a package for which the two ratings are higher than the package that has been accepted.
Inspire has a post-settlement stage, during which it uses the preference information provided by each user to determine whether it is possible to construct packages that are better for the two parties. In this negotiation Inspire determined that Misty's and Smiley's settlement could be improved. In this simple negotiation there is only one such package:
320 000 $
Misty asked Smiley about this new offer and after a short exchange they both agreed that this last package is superior to the compromise package. Thus the negotiation is completed.
The complete graph of the history of negotiation shows the effect of accepting this alternative. To show the differences between Misty's and Smiley's view points we present both graphs noting again that each negotiator can see only his or her graph and cannot see the opponent's graph.
The graph that Misty sees:
The graph that Smiley sees: