Robots will make better negotiators

Negotiation is one of the oldest human activities and is an important part of our economy.  It is essential for sharing resources between people and groups.  However, as our organisations have become more complex, the outcomes that we are achieving are getting worse not better.

The usual objective of negotiation is to match the needs of two parties.  In business, this is often bringing together the provider of goods or services with someone who is in need of the resource or to get the holder of budget to release it for a given initiative.

Negotiations have three potential outcomes: win-win, win-lose or lose-lose.  Amazingly, talking to many businesses over many years, I’ve come to the conclusion that lose-lose is increasingly the most common outcome.  That is because in the absence of confidence, people opt to avoid loss and choose not to act.  But in choosing not to act, they are adding friction into their businesses and missing out on the return on risk they should be taking.

As business has become more complex, so it has become much harder for buyers to know what a good outcome is.  Worse, this is becoming a major cause of stress in business as both buyers and sellers have a huge amount at stake and lessening ability to navigate to a good result.

Consumer example

To illustrate the problem, consider this case we can all relate to.  You are in the market for a car, but have limited time and need to buy now.  You know the car make and model you want and know the recommended retail price.  You walk into a dealer who makes an on-the-spot offer for the car that matches your requirement at 10% less than the listed price.  Do you sign?

The dealer could be offering you a great deal, or they could be holding out and most customers could be getting 15% off.  You just don’t know.  Meanwhile, if it is the former the dealer is frustrated and has less incentive to offer a good starting price for future customers.

Independent brokers are sometimes a solution for this dilemma, having knowledge of the market and knowing which price is actually a fair one.

In business

Few executives have any more time available than our hypothetical car buyer.  Also like the car buyer, most things they are seeking to acquire are outside of their day-to-day experience.

It’s relatively easy when commodity items (like replacement parts for machinery) are needed, procurement experts can negotiate against a pricelist.  However, when it’s a complex product, such as a computer system or an internal budget allocation for a new service, then there are few points of reference within the organisation.

I lamented in a past post about the lack of productivity growth resulting from our transition to a digital economy (see Where is the digital-fuelled growth?).  Where new approaches to sharing knowledge takes friction out of the systems, there is a boost to productivity and this is where much of our effort should go.  Negotiations are a prime candidate for this focus.

Increasingly digital solutions are allowing for the creation of anonymous or semi-anonymous benchmarks.  But complex procurement and negotiations require more than simply finding a fair price.  Factors at play often include risk, time, quality, the competitive landscape and so many more.

Artificial Intelligence and Robotics

The two technologies that could make a material change here are Artificial Intelligence and Robotic Process Automation.

Artificial Intelligence, or Cognitive Computing, is a form of very advanced analytics.  In negotiations between parties, the primary objective is for the person who wants to acquire the resource to work out what a fair trade would look like.  Even when it is just the allocation of internal budget for a new capability, there is still a need to know whether the return is commensurate.

Where these technologies come in is their capability to find things that are similar based on a wide range of criteria.  For example, anyone who has seen how well search engines can group similar questions, worded completely differently, has some idea of how Cognitive Computing can bring together the right answers from disparate sources.

Where Robotic Process Automation (RPA) can be most effective is in taking the emotion out of negotiation and pushing to get the best outcome based on relative criteria.  There is an argument that the political and emotional process is an important part of getting to the best possible outcome.  The problem with this is that as the environment has become more complex, negotiations have become simplified on a subset of dimensions meaning that it isn’t the best argument that wins but, all too often, no argument that wins.

Most negotiations are a manual process with lots of spreadsheets and lists of points to push on and others to give on when pushed.  This is exactly the sort of complex process that RPA is ideal at and makes it an ideal target for the technology.

Win-win

Our robots can be programmed to provide a win for us all.  While some negotiators are happiest with a win-lose, win-win puts the right incentives in the system for the long-term.  If we program our robots that way we can be freed-up for the creative task of getting the right job done with the right resources.

In a future economy, where robots act on our behalf to find winning combinations, everyone could win.