Strategy dynamics

The word ‘dynamics’ appears frequently in discussions and writing about strategy, and is used in two distinct, though equally important senses.

The dynamics of strategy and performance concerns the ‘content’ of strategy – initiatives, choices, policies and decisions adopted in an attempt to improve performance, and the results that arise from these managerial behaviors.The dynamic model of the strategy process is a way of understanding how strategic actions occur. It recognizes that strategic planning is dynamic, that is, strategy-making involves a complex pattern of actions and reactions. It is partially planned and partially unplanned.

A literature search shows the first of these senses to be both the earliest and most widely used meaning of ‘strategy dynamics’, though that is not to diminish the importance of the dynamic view of the strategy process.

Static Models of Strategy and Performance

The static assessment of strategy and performance, and its tools and frameworks dominate research, textbooks and practice in the field. They stem from a presumption dating back to before the 1980s that market and industry conditions determine how firms in a sector perform on average, and the scope for any firm to do better or worse than that average. E.g. the airline industry is notoriously unprofitable, but some firms are spectacularly profitable exceptions.

The ‘industry forces’ paradigm was established most firmly by Michael Porter, (1980) in his seminal book ‘Competitive Strategy’, the ideas of which still form the basis of strategy analysis in many consulting firms and investment companies.
Richard Rumelt (1991) was amongst the first to challenge this presumption of the power of ‘industry forces’, and it has since become well-understood that business factors are more important drivers of performance than are industry factors – in essence, this means you can do well in difficult industries, and struggle in industries where others do well. Although the relative importance of industry factors and firm-specific factors continues to be researched, the debate is now essentially over – management of strategy matters.

The increasing interest in how some businesses in an industry perform better than others led to the emergence of the ‘resource based view’ {RBV) of strategy (Wernerfelt, 1984; Barney, 1991, Grant 1991), which seeks to discover the firm-specific sources of superior performance – a research interest that has increasingly come to dominate research in

The need for a Dynamic Model of Strategy and Performance

The debate about the relative influence of industry and business factors on performance, and the RBV-based explanations for superior performance both, however, pass over a more serious problem. This concerns exactly what the ‘performance’ "is" that management seeks to improve. Would you prefer, for example, (A) to make $15m per year indefinitely, or (B) $12m this year, increasing by 20% a year, starting with the same resources?

Nearly half a century ago, Edith Penrose (1959) pointed out that superior profitability (e.g. return on sales or return on assets) was neither interesting to investors – who value the prospect of increasing future cash flows – nor sustainable over time. Profitability is not entirely unimportant – it does after all provide the investment in new resources to enable growth to occur. More recently, Rugman and Verbeke (2002) have reviewed the implications of this observation for research in strategy. Richard Rumelt (2007) has again raised the importance of making progress with the issue of strategy dynamics, describing it as still ‘the next frontier … underresearched, underwritten about, and underunderstood’.

The essential problem is that tools explaining why firm A performs better than firm B at a point in time are unlikely to explain why firm B is growing its performance more rapidly than firm A.

This is not just of theoretical concern, but matters to executives too – efforts by the management of firm B to match A’s profitability could well destroy its ability to grow profits, for example. A further practical problem is that many of the static frameworks do not provide sufficiently fine-grained guidance on strategy to help raise performance. For example, an investigation that identifies an attractive opportunity to serve a specific market segment with specific products or services, delivered in a particular way is unlikely to yield fundamentally different answers from one year to the next. Yet strategic management has much to do from month to month to ensure the business system develops strongly so as to take that opportunity quickly and safely.What is needed, is a set of tools that explain how performance changes over time, and how to improve its future trajectory – i.e. a dynamic model of strategy and performance.

A Possible Dynamic Model of Strategy and Performance

To develop a dynamic model of strategy and performance requires components that explain how factors change over time. Most of the relationships on which business analysis are based describe relationships that are static and stable over time. For example, “profits = revenue minus costs”, or “market share = our sales divided by total market size” are relationships that always were true and always will be. Static strategy tools seek to solve the strategy problem by extending this set of stable relationships, e.g. “profitability = some complex function of product development capability”.Since a company’s sales clearly change over time, there must be something further back up the causal chain that makes this happen. One such item is ‘customers’ – if the firm has more customers now than last month, then (everything else being equal), it will have more sales and profits.

The number of ‘Customers’ at any time, however, cannot be calculated from anything else. It is one example of a factor with a unique characteristic, known as an ‘asset-stock’. This critical feature is that it accumulates over time, so “customers today = customers yesterday +/- customers won and lost”. This is not a theory or statistical observation, but is axiomatic of the way the world works. Other examples include cash (changed by cash-in and cash-out-flows), staff (changed by hiring and attrition), capacity, product range and dealers. Many intangible factors behave in the same way, e.g. reputation and staff skills.Dierickx and Cool (1989) point out that this causes serious problems for explaining performance over time:
* "Time compression diseconomies" i.e. it takes time to accumulate resources.
* "Asset Mass Efficiencies" ‘the more you have, the faster you can get more’.
* "Interconnectedness of Asset Stocks" .. building one resource depends on other resources already in place.
* "Asset erosion" .. tangible and intangible assets alike deteriorate unless effort and expenditure are committed to maintaining them
* "Causal ambiguity" .. it can be hard to work out, even for the firm who owns a resource, why exactly it accumulates and depletes at the rate it does.

The consequences of these features is that relationships in a business system are highly non-linear. Statistical analysis will not, then, be able meaningfully to confirm any causal explanation for the number of customers at any moment in time. If that is true then statistical analysis also cannot say anything useful about any performance that depends on customers or on other accumulating asset-stocks – which is always the case.

Fortunately, a method known as system dynamics captures both the math of asset-stock accumulation (i.e. resource- and capability-building), and the interdependence between these components (Forrester, 1961; Sterman, 2000).The asset-stocks relevant to strategy performance are resources [things we have] and capabilities [things we are good at doing] . This makes it possible to connect back to the resource-based view, though with one modification. RBV asserts that any resource which is clearly identifiable, and can easily be acquired or built, cannot be a source of competitive advantage, so only resources or capabilities that are valuable, rare, hard to imitate or buy, and embedded in the organization [the ‘VRIO’ criteria] can be relevant to explaining performance, for example reputation or product development capability. Yet day-to-day performance must reflect the simple, tangible resources such as customers, capacity and cash. VRIO resources may be important also, but it is not possible to trace a causal path from reputation or product development capability to performance outcomes without going via the tangible resources of customers and cash.

Warren (2002),(2007) brought together the specification of resources [tangible and intangible] and capabilities with the math of system dynamics to assemble a framework for strategy dynamics and performance with the following elements:
* Performance, "P", at time "t" is a function of the quantity of resources "R₁" to "R_n", discretionary management choices, "M", and exogenous factors, "E", at that time (Equation 1).(1) "P"("t") = f{"R₁("t")", .. "R_n("t")", "M"("t"), "E"("t")}