Applying game theory to strategy practice
Game theory is a tool for analysing strategic interactions that help improve the quality of decision-making in any organisation
- Article |
- 01 March, 2012
- Although there are many potential applications of game theory in business, most companies fail to appreciate its practical usefulness in strategic decision making.
- Game theory is a powerful tool for analysing strategic interactions where decisions are interdependent.
- By identifying the right players and understanding their motivations, managers can generate the right options to set up the game and choose the optimum strategy.
“India’s Bajaj Auto Ltd will set up a joint venture (JV) with the Renault-Nissan alliance to produce and market small cars — but priced higher than Tata Nano, the companies announced on Monday. “For Bajaj Auto, the project is important as Tata Nano will be a serious threat to its three-wheeler business”.
“India’s leading retail chains such as Future Group, Reliance Retail, Shoppers Stop, Style Spa and Woodland are advancing their annual discounts by up to three weeks, facing a huge slump in sales since Diwali. “Our inventory situation is not bad, but we also started our sales since the entire market is getting into discount mode”, says Spencer’s retail executive director Subrata Siddhanta”.
The Bajaj story transpired a mere two days before the launch of the Tata Nano during Auto Expo 2008 in Delhi. It signalled Bajaj’s intention to transition from an established bike manufacturer to a passenger-car maker. The second case of Indian retailers advancing their annual discounts is more recent and shows how the movement of one market player affects choices made by others. Both news stories provide evidence of science at work in the making of strategic decisions.
When looked at closely, these and numerous other examples show the potential for applying game theory to strategic decisions. However, managers are yet to fully assimilate the scientific rigour and discipline of game theory to the art of strategic decision making. Perhaps they are averse to the idea of applying theory to business, which many believe to be a more ‘practical’ field. They may also lack the time or interest to do the rigorous analysis that game theory demands. In the case of big decisions, such detailed analysis is necessary to highlight all the possible options available, some of which, under traditional strategic analysis, would be unthinkable.
The applicability of game theory to economics first came to light when John von Neumann and Oscar Morgenstern published their path-breaking book Theory of Games and Economic Behaviour in 1944. The book showed how a scientific approach can be applied to making decisions influenced by decisions made by other players. However, the influence of game theory did not become widespread until 1994, when three economists: John C Harsanyi, John F Nash and Reinhard Selten were awarded the Nobel prize for their pioneering work on non-cooperative games.
Game theory is a tool for analysing strategic interactions. It helps improve the quality of strategic decisions in situations where decisions are interdependent. In oligopolistic situations where a few firms dominate the market - the prevailing condition in almost all industries - strategic decisions have an effect on other players in the market. Any move by one company is closely followed by their competitors. Therefore, when one takes into account how other players are going to react to the boardroom metrics, action is taken. The optimum choice may lie somewhere between the best and the worst choices seen in isolation. These decisions are based on reasoning backward and thinking forward, so we are able to understand the perspective of others in order to choose the best path, considering other players are also going to choose the best course of action for themselves.
Applying the theory
While game theory has many applications in economics, policy development and business negotiations, this article focuses on its use in the making of strategic choices by managers, when their decision will be affected by the actions of other players in the market. Its application potential ranges from decisions about a capacity addition to market entry, and from pricing to new product launches.
Capacity addition is one of the key strategic decisions managers have to make. Deciding how much capacity to add and when to add it has a bearing on industry profitability. In such cases, adding capacity draws strong reactions from competitors. As long as supply lags market demand, incumbents can earn an economic rent and generate supernormal profits. Any challenge to this situation can generate strong reactions which need to be accounted for when taking a decision to add capacity.
The decision to add capacity can be modelled by either considering two or three choices, depending on their situation. Companies can decide to add or not to add, or make no addition, a small addition or a large addition. While modelling the game, it is possible to consider simultaneous moves; where one player does not know what others are going to do, and sequential moves; where one player has increased their capacity and competitors must decide whether to follow.
Market entry is another important area where game theory can be extremely useful in improving decision making. Understanding how incumbents would behave following the entry of a new player provides critical insights in deciding; whether to enter the market or not, when is the correct time to do so, and how one can enter in the best possible way.
If the market is new for all the players, game theory can help in deciding whether the first mover will have an advantage. In some cases, it may be worth waiting for others to test the new market in order to learn from their mistakes. In market entry decisions, there is also a possibility of changing the game being played currently. In mobile telephony, companies were playing the high-priced game until Reliance Infocom arrived on the scene. They first bundled a handset with a mobile connection and then launched the ‘Monsoon Dhamaka offer’ with call rates at just 40 paisa per minute and a bundled handset at `501. The entry strategy of Reliance Infocom changed the mobile telephony game in India forever.
Pricing is a very potent weapon for managers if used correctly. Competitors closely observe price movements, particularly in oligopolistic markets. Game theory can help when modelling the impact of pricing decisions on competitors and assist in deciding on the best choice available, given the most likely response of these competitors. In pricing decisions, some players may emerge as leaders in the market. They first decide when, and by how much to change the price, while others follow them in a way which is aligned with their own market position and results in the best payoffs, given the move made by the leader.
New product launch
We have come across a number of cases where companies launch new products with great fanfare while ignoring or miscalculating the moves of other players in the market. The demand projections are often based on the industry’s average growth rate, a rate which may not apply to all players in the same way. Moreover, the response of other players to new product launches may alter projections drastically if managers do not account for their motives while making these projections. By understanding which players will be affected by the new product and how they will react, companies can position their launch in a way that achieves their business goals.
Setting up and playing the game
As can be seen from the above applications, managers can make better strategic decisions by applying game theory models. They should take five steps to apply game theory in an effective way to improve the quality and timing of their decisions.
The first step in setting up the game for strategic interactions lies in the identification of the relevant players. Every strategic move of the company may not be relevant for all the players and hence may not provoke reactions by them. For example, if it is a competitive interaction, not all competitors may react to a decision. The concept of strategic grouping can be applied here to identify which players will react to a given decision. Within an industry, competitors fall into strategic groups where they have close competition within the group, but no direct competition outside the group. For example, in the automotive industry SUV manufacturers will fall into one strategic group while small car makers will be in a different strategic group.
The motivation of players also varies widely. Some players may be in the market to protect their existing business, while others may be there to leverage their customer relationships. Both of them may, or may not be ‘serious’ players, as we found during our client work in the solar energy industry (see exhibit 1).
Sometimes there is a need to look at market players beyond the competitors and include suppliers and customers. This is because their interests may be affected by the company’s decisions, leading them to respond either positively or negatively to the new decision. For example, if a company decides to launch an ultra-low-cost product, how will its suppliers respond to this? If the decision may hurt their interests or require them to make new investments, their response to the new product development needs to be accounted for.
Get into the heads of other players
Once we know who the relevant players are, it is very important to understand their behaviour and motivations. To understand how these players will act and react to various moves, one has to get into their heads. Mapping past competitive interactions can help in understanding how different players respond to various strategic moves. However, they will not necessarily behave in the same way in the future. Their strategic direction might have recently changed, requiring them to respond more aggressively or more moderately. For example, if a multi-business corporation has increased emphasis on the business unit where you are competing, they may now be more aggressive in their strategic conduct. It is also important to profile the decision makers at various levels within competitor organisations. If there is a major change in the top team, it is likely that the organisational structure of the company may change. Profiling decision makers can help predict rival organisations’ behaviour more accurately.
Build a matrix of their available options
Identifying relevant players and understanding how they will behave under different conditions is just the starting point. Managers then need to lay out the various options available to them in order to set up the game. This step can be best illustrated by our client work in the solar energy industry. The client had developed an advanced product which was superior to existing products in the market — both their own products as well as their competitors’. Pricing the new product was a crucial decision prior to its launch. In making this decision the client had three options available to them:
- Price the new product above the price of the existing products and skim the market
- Price the new product at the same price as the existing products (and price existing products at a lower price) to increase market share
- Price the new product below the price of the existing products (and price existing products at a much lower price) and ‘kill the competition’
The client could choose any of these options. The best option would depend on understanding the options available to competitors and which one they were likely to choose. For any option chosen by the client, competitors also had three options: do nothing, reduce prices and match warranty. The competitors could do nothing. They could also reduce their prices to make their offering look attractive to customers. Competitors could also match the warranty of the new product — which was the key selling point here — to make their product ‘comparable’ in value to the new product. In this way, the game could be set up as a sequential game, where the client had three options and for each one of their competitors also had three options.
Assess the feasibility and payoffs
The game tree, as shown in exhibit 2 above, provides all the possible options that players had at their disposal. However, before these options were shortlisted there was a need to assess their feasibility. For example, the competitors had an option to launch a similar product. However, it would have taken them at least six months to import and two years to manufacture. Therefore, those options were not considered as feasible at the time of the client’s product launch. However, the client was advised to monitor the competitors’ activities and set up the game differently as soon as a competitor began launching a similar product.
Within the feasible options, there is a need to estimate the payoff of each choice to one’s self and to one’s competitors. The payoffs can be estimated in terms of profitability, market share or any other indicator which is relevant to the decision and is aligned with the company’s overall objectives. In some cases, it may be difficult and less practical to quantify the payoffs because of imperfect information, especially when the companies are privately owned. In such cases, it will still be worthwhile to play the game qualitatively, that is, where the outcome can be simply winning or losing, without any specific value being attached to the outcome.
Choose the optimum path
The optimum path will then depend on what the best scenario is when all players act rationally to maximise their payoffs. While the chosen path may not be optimum to maximise collective gains, the decision is taken because the other player will also try to maximise their payoff, the result being an outcome that neither party would favour.
Game theory has emerged as a vital tool in the field of economics and business for developing better strategies when the fate of one player depends on the decisions made by others. By identifying the right players and understanding their motivations, managers can generate the right options to set up the game and choose the optimum strategy.