The Game Theory of Procurement and RFPs

Jeff Bezos is one of the wealthiest men in the world. He is extraordinarily good at management. He has a useful analogy for decisions .

“Some decisions are consequential and irreversible or nearly irreversible – one-way doors – and these decisions must be made methodically, carefully, slowly, with great deliberation and consultation. If you walk through and don’t like what you see on the other side, you can’t get back to where you were before. We can call these Type 1 decisions. But most decisions aren’t like that – they are changeable, reversible – they’re two-way doors. If you’ve made a suboptimal Type 2 decision, you don’t have to live with the consequences for that long. You can reopen the door and go back through. Type 2 decisions can and should be made quickly by high judgment individuals or small groups.”

For a Type 2 decision in a procurement context, we can imagine executing a pilot of a software program. It will cost much less than what a license or subscription would cost. The buyer can test it out and see how it works, how responsive the support is, how willing the vendor is to make changes based on the experience, etc. If the pilot project doesn’t succeed, then no harm, no foul. Uninstall the software or stop using it and move on. You’ve picked up inexpensively some valuable information that may or may not come in handy in the future.

A Type 1 decision in the Bezos framework involves purchasing something that the buyer is going to live with for a long time, involving significant expense. For example, it could involve purchasing capital equipment for a new semiconductor manufacturing facility. This choice will determine the kinds of chips the fab can produce. Will these be in demand? For how long? Will the facility be able to earn a high enough price with a sufficient volume to generate at least the required rate of return? How vulnerable is this type of chip to the cyclicality that stalks semiconductor markets? This needs to be a thoughtful decision.

In his letter, Bezos goes on to descry the natural tendency to bureaucratize everything by approaching too many things as a Type 1 decision.

“As organizations get larger, there seems to be a tendency to use the heavy-weight Type 1 decision-making process on most decisions, including many Type 2 decisions. The end result of this is slowness, unthoughtful risk aversion, failure to experiment sufficiently, and consequently diminished invention. We’ll have to figure out how to fight that tendency.”

As Farnam Street points out, it’s not a dichotomy between reversible and irreversible; it’s a continuum. You can always get out of a decision. There is a cost to be paid, but you always have an exit. The higher the cost, the more difficult the decision is to reverse, putting it closer to the irreversible end of the spectrum.

So, the rational thing to do is to adapt your process to the type of decision.

“Make reversible decisions as soon as possible and make irreversible decisions as late as possible.

“When decisions are reversible, make them fast.?Your biggest risk is dragging your feet and not making a decision. The cost to acquire additional information isn’t worth the effort.

“When decisions are irreversible, slow them down.?The biggest risk is making the wrong decision. The cost to get the information we need to reduce uncertainty is worth the time and effort.”

To try and map how this logic might work in procurement, let’s put on our game theory hat.

Imagine a game in which we have to guess the picture underlying each of a series of ten puzzles. Every player gets a different set of puzzles, some of which may overlap partially across players.

Let’s assume that in this game we get very few points for guessing the picture for small puzzles correctly and we get a tremendous number of points for identifying the picture of large puzzles.

The more pieces of an individual puzzle we have, the easier it is for us to determine what its picture may be.

According to the rules of the game, we have a budget of time and money. With the budget, we can purchase additional pieces for any puzzle. We don’t have to purchase pieces, but we can.

Players can only make one guess. So they wait until they’re comfortable.

The winner of the game is the player who can identify correctly the greatest number of puzzle pictures in a given period. We’ll break ties by comparing the amount of remaining money each tied player has left, with the individual who finishes the game with the most money winning.

Some players will have an advantage because they may have played one or more of the puzzles previously. They’ll be able to recognize pieces faster and they can skip ahead, moving from one to the next.

What is the optimal strategy in this game?

You would spend most of your time and money on the big puzzles. The small puzzles aren’t worth more than a cursory effort.? For these, you can make an educated guess and move on. With respect to the larger puzzles, she can try and purchase more pieces, conscious of the trade-off. This is where judgment comes into the picture.

Let’s throw four twists into the game.

Twist #1: you can talk to other players and ask them questions. Since everyone is playing different sets of puzzles, it might make sense to collaborate. Player A can help Player B and Player B can reciprocate potentially on a subsequent puzzle. There is no cost to cooperation.

Twist #2: you have a private database of every puzzle you have played in the past. It is easy to search through. The more puzzles you have played in the past, the more likely you are to find one in your history that you are now being asked to solve.

Twist #3: everyone can access a community database of certain puzzles that others before you have seen.

Twist #4: you can pool your efforts with other players if you see that they are working on what appears to be the same puzzle as you.

These twists now change the strategy a rational player should employ.

In addition to the basic strategy of focusing on the larger puzzles, the player should scan the databases for signs of the current puzzle under examination. When it comes to the larger puzzles, if she is making progress, then she should continue. If she hits a roadblock, instead of switching puzzles, she can ask for advice or seek collaboration. These preserve budget and they speed up the time.

Naturally, the more experienced players will have seen more previous puzzles and will be more adept at using the databases. But anyone can employ collaboration, even (and especially) the novices. The collaboration levels the playing field potentially more so than the databases which require familiarity.

How does this game translate to procurement?

Procurement is like trying to solve a puzzle. The buyer wants to have as many pieces as possible to identify the correct picture. However, too often they don’t see enough competition on price and solution and they end up taking a stab in the dark.

It’s best to focus the most procurement department resources on the larger purchases.

Ideally, we’d have structured data of prior procurements: the RFPs the buyer issued, the proposals they received, the vendor they selected, and the contract they signed, in addition to structured data on procurements in the public sector.

It’s best to get information from other buyers. And, if possible, to purchase jointly with other buyers who want to purchase the same products we do. So there should be some sort of networking and collaboration tools built into the process.

This is what we have built at EdgeworthBox . We have a set of tools, structured data (in public and private repositories), and community to help B2B buyers purchase the right solution, from the right supplier, at the right price. We are changing the game to make it easier and more efficient. Give us a shout. We’d love to hear from you.