FITNESS LANDSCAPES

Let's borrow the topological analogy of fitness landscapes from Stuart Kauffman, a MacArthur Fellow and one of the brilliant minds at play at the Santa Fe Institute. To suit our purposes, let's apply his analogy of landscapes to spaces of possibility to be explored. The optimum position on any landscape would be defined as the highest "peak" and the least attractive position would be considered the lowest "valley". What distinguishes a peak as an optimum point in the landscape is that a minimum amount of resources is required to obtain a maximum amount of sustenance (profit). The summit of a peak is reached when nothing can be improved in terms of our survival. Life is easiest and most rewarding at a peak.

Let's approach the analogy of fitness landscapes from two different perspectives, then combine these two perspectives, and see what the combination has to offer in terms of strategy for an enterprise. Our first application of the analogy will be to assess fitness for survival and sustenance. The next application will be to assess strategy and future. The combination of these two perspectives mirrors the corporate requirements for current production and development of future possibility.

The nature of peaks, in any competitive environment, is that they are in a state of constant flux; the same applies to fitness landscapes. Peaks in both situations vary over time as well as, space. As a landscape changes, any particular locale will either increase or decrease in its fitness for an entity's survival. The counter of this is also true; the fitness of any entity, for survival at any particular locale on a landscape, also changes. Peaks of fitness are optimal and valleys of fitness are the least advantageous. At a peak, the greatest returns result from minimal output of energy. In a valley, very hard work produces a meager existence and survival is constantly in question.

There are a variety of landscapes we may encounter. One contour is of a single peak with relatively steep sides. On this Fujiyama-like landscape, every step will move us toward or away from the peak, and what direction we are taking with each step will be immediately obvious to us. Most corporations view their surrounding landscape in this way, and the way that most of them choose to express their strategic intent are an indication of this. "Being number one" points to the perception that there is a single peak, and that whoever is perched on top is king.

Another landscape contour is of two or a few peaks of relatively equal height. This way of interpreting our environment offers essentially the same challenges as a single-peak landscape because what sets precedent is reaching a peak, and each step is measured in terms of climbing toward a more optimum point. If a corporation is perceiving the landscape in this way, the participants of that corporation will have a shared intention of focusing solely on winning the game that is already underway. The feature of predictability and ability to know what is wanted is similar in this view of the competitive landscape to the previous one.

Landscapes can also be flat. In this type of terrain, it doesn't matter whether or not you explore because any point on the landscape is equally suitable for survival, and no advantage can be gained by adaptation. Because most bureacratic businesses operate in this way, they manifest as tired and sleepy enterprises with little concern for intention, change, or challenge. Low investment of resources, time and attention into searching for new possibilities characterises these companies.

These three might be considered landscapes of predictability. That is, their shape and characteristics are largely known, or believed to be known, and one's position in the landscape can be seen relative to the whole.

The fourth, a rugged landscape or a landscape of possibility, which has the most to offer for developing corporate strategy. These landscapes have many peaks of varying heights, and of the many peaks, some are more outstanding than others. Between each of the peaks are valleys. What's interesting about this landscape is that, even though it's obvious that we are either climbing or descending, we cannot tell whether we are climbing towards one of the optimum peaks on the landscape, or one of the lowest peaks. The efforts of our climb may be poorly rewarded because other peaks are higher, and we would have produced better results with our efforts had we chosen to climb another peak. In other words, we may win in the wrong marketplace and thus lose.

Again, Stu Kauffman has created a useful distinction for referring to the exploration of this fourth type of landscape -- he calls it an adaptive walk. Namely, as we explore the surrounding landscape, we assess our fitness level at each point and either adapt our capacities to increase fitness at that point, or move in a direction in which our existing capacities would be more effective. We improve our position on the landscape by either adapting internally (which improves local efficiency), or by changing our location on the landscape. A landscape of possibility includes all areas not yet explored, most of which cannot be seen from one's current position -- or from any other single position.

The nature of a rugged (rapidly changing) fitness landscape is that we cannot see beyond our immediate vicinity in order to predict the long-term effects of our actions. It can be likened to walking in a jungle. Even if we climb a tree and are able to see that the mountain we are seeking is in a certain direction, we still are not able to see the best pathway to get to the mountain from that vantage point. Or we may just happen to be in a position in which the next step, which is hidden from view, will plunge us into a deep pit. Our walk in a jungle may put us at a vantage point from which we may not see the impassable river ahead; one that will require we go in the opposite direction from our destination for some considerable length of time.

These same principles apply to the concept of a fitness landscape. Namely, from a specific vantage point, we can see the direction to a higher peak but we cannot see the pathway to it. Nor can we tell if it will prove to be a particularly high peak, when seen within the context of a larger landscape, until we actually get to the top of that peak. From our new vantage point, perched atop the peak, we may see that our current peak was obscuring numerous higher peaks that would have been more beneficial to pursue from the valley floor.

Once we've reached a fitness peak in a landscape of possibility, there are no immediate moves that can be made that improve our effectiveness. As we climb, the higher we get, the less likelihood there is of any benefit from any particular step. If we are only part way into a climb, any step will have the possibility of improving or eroding our fitness for survival. If we are still in a valley, any move will better our position, but as we move away from a valley floor and toward a peak, each step reduces the possibility for improvement in the following step, and increases the possibility for a reduction in effectiveness.

As Cray developed its ability to build supercomputers, it became locked into climbing that particular peak. The better it got, the more expensive it became to pursue any other path - such as developing minicomputers. As IBM developed leasing arrangements for selling computers, it became increasingly difficult to adapt to a marketplace with other financial options. Xerox is fabled for having invented many things which were exploited by others successfully. Those things did not advance Xerox on the particular peak that it had created and climbed. While many consider this a failing of Xerox, it may represent its strength when seen from the perspective of exploring rugged landscapes.

The challenge of our adaptive walk is intensified by the fact that most of the peaks we are able to see are merely local peaks and may in fact be low when seen within a larger landscape. Arriving at a local peak, even though it is low, may be sufficient for success, if entities of higher fitness from other peaks are excluded from our area. In many cases, however, higher peaks have attained their height on competitive landscapes by encouraging their occupants to explore lower territories on numerous other peaks. Within a longer time frame on competitive landscapes, those that have conquered higher peaks can invade the territory of those lower on the landscape -- but frequently this does not work the other way around.

As the number of peaks in a landscape increases, the distance between individual peaks lessens and the number of steps involved in reaching the top of a peak decreases. As we move closer to the top of each peak, there tends to be an increase in the number of peaks that are not optimal for us. Within real life situations, the most common condition that occurs is that there are substantially more low peaks than high ones. There is no way of knowing which peak is optimum because there is no way of searching the entire landscape. What proves to be an invaluable asset for exploring landscapes of possibility is a capacity for climbing, combined with a capacity for searching out peaks that are higher than the current one being climbed.

As we move uphill, the number of steps that look like they produce benefit declines and it becomes compelling to climb the particular hill that we are on and narrow the number of directions by which we approach that peak.

As a company begins to dominate a market, the number of options open to it lessens. Also new limitations are introduced as to what specific pathway can be taken to the top. A company's interest and ability to look for new peaks is discouraged as the apparent cost of exploration increases. The increase in exploration costs, created by expending energy on the lower part of a peak are, of course, deducted from our profits. This is another way of saying, we're not interested in investing in our future at the expense of our current advantages.

The theories of fitness landscapes are applicable to technology, production efficiency, marketing approaches, specific products, features of products, and knowledge bases. And for today's rapidly changing world, these theories offer countless insights for working with organizational designs. There are a variety of organizational forms, each having its own unique qualities of fitness. More frequently than not, today's corporate winners are possessors of unique structures of organization which have increased capacities for adaptive walks.

Entire ranges of peaks can be dominated simply by capturing a single peak. Even larger ranges can be dominated by conquering a number of related peaks. By increasing complexity by integrating a number of peaks, a company can greatly expand its ability to dominate ranges of peaks and its ability to create as well as explore new peaks.

By being fit enough to position itself atop a particular peak in operating systems, Microsoft dominated IBM, and now IBM must clone. Then by continuing to hold that position, Microsoft was able to dominate a number of smaller peaks in the arena of software applications. This is enabling access to peaks in banking, information use, and entertainment. By combining these as well, and thereby increasing its complexity - its various connections, couplings and relationships - Microsoft is able to provide a more robust ability to explore landscapes. Equally important, it increases its natural adaptation to changing landscapes. By the time its domination of operating systems is over - that day will come - it may have both established itself on new peaks and created a first class ability to explore new landscapes of possibility.