Routines, heuristics, and innovation

Both in customary usage and in our technical use of the term, “inno­ vation” involves change in routine. We have stressed the uncertainty that inevitably surrounds technical innovation-the implementation of a design for a new product, or of a new way to produce a product. A similar uncertainty surrounds other kinds of innovation-the es­ tablishment of a new marketing policy, or a new decision rule for restocking inventories . In general, two kinds of uncertainty sur­ round these innovations. The precise nature of the innovation actu-ally arrived at is usually not closely predictable at the start of the endeavor that culminates in the innovation. And the consequences of  employing  the  innovation- changing  the  routine-in  general will not be closely predictable until a reasonable amount of actual operating experience with it has been accumulated. There is, how­ ever, more to be said about the relations of routine behavior and in­ novation than to observe that these concepts are commonly (and appropriately) regarded as opposed ideas. Our final task in this chapter is to explore some of the subtler connections between routin­ ization and innovation, and ultimately to indicate how the existence of innovative activity relates, in our evolutionary theory, to the gen­ eral image of firm behavior as governed by routine.

1. Puzzles from Prevailing Routines

It is sometimes remarked of an important research achievement that the hard part was in locating the right question; finding the answer to that question then proved to be relatively easy. One way in which the routine functioning of an organization can contribute to the emergence of innovation is that useful questions arise in the form of puzzles or anomalies relating to prevailing routines. The con­ creteness of such questions and the obvious existence of an applica­ tion for the answers is an important point in their favor as guides to problem-solving activity.

Consider the foreman of a work team responsible for a particular operation (set of routines) who observes that a machine is not work­ ing properly. He routinely calls in to the maintenance department, which in turn routinely sends out a machine repairman. The ma­ chine repairman has been trained to diagnose in a particular way the troubles that such a machine might have. He goes down a list of pos­ sible problems systematically, and finds one that fits the symptoms.

. He fixes the part so that the machine again may play its role in the overall work routine .  He may also, however, report to the foreman that this particular kind of trouble has become very common since the supplier started using aluminum in making the part in question and that perhaps the machine should be operated in a different manner to  avoid the difficulty .

Or consider a s ales manager who observes a significant and sus­ tained decrease in total sales of a particular item. He routinely calls in his young assistant- a recent graduate of a master’s program in management- to do a study of the problem. The assistant, with a bit of clerical help, scans what has been happening to sales in particular regions and by particular salesmen. He ascertains that almost all of the decrease has occurred in the Southeast. He may go on to check up on the activities of the salespeople concerned with the Southeast and may recommend some replacement of personnel . He may suspect that some important change in demand conditions has occurred and propose a new market survey to discover its nature. Or he may pro­ pose that a new advertising campaign, addressed to customers in the Southeast, may be needed.

These examples illustrate, on the one hand, the routine func­ tioning of organizations. The responses described fall into the typical pattern in which a crisis or “exception” condition in one part of the organization is part of the routine content of jobs of other personnel. On the other hand, it is significant that the problem-s olving responses routinely evoked by difficulties with existing routines may yield results that lead to major change. The effort triggered by the re­ pairman’s suggestion may lead to a radical improvement in the method of operation of the machine, or to a decision to switch to ma­ chines of quite a different sort, req uiring numerous adaptations else­ where in the routine. The market survey proposed by the young as­ sistant may indicate that the trouble in the Southeast is only a symptom of a market change that is likely to become pervasive, and may thus trigger redesign of the product to meet the specific chal­ lenge that the survey i dentified. Problem- solving efforts that are ini­ tiated with the existing routine as a target may lead to innovation in­ stead .

2. Existing Routines as Components

Schumpeter identified innovation with the “carrying out of new combinations” (Schumpeter, 1934, pp. 65 -66) . This phrase gives useful emphasis to the fact that innovation in the economic system-and indeed the creation of any sort of novelty in art, sci­ ence, or practical life-consists to a substantial extent of a recombin­ ation of conceptual and physical materials that were previously in existence. The vast momentum of scientific, technological, and eco­ nomic progress in the modern world derives largely from the fact that each new achievement is not merely the answer to a particular problem, but also a new item in the vast storehouse of components that are available for use, in Unew combinations,” in the solution of other problems in the future.

Innovations in organizational routine similarly consist, in large part, of new combinations of existing routines. An innovation may involve nothing more than the establishment of new patterns of in­ formation and material flows among existing subroutines. It may in­ volve the replacement of an eXisting subroutine by a new and dif-ferent one that performs, in relation to the rest, the same function that the old one did. Some parts of the innovative routine may rely on physical principles only recently discovered and now imple­ mented through novel types of equipment and newly developed skills-but surrounding this novel core there may be many layers of complementary activity governed by the same routines that have prevailed for many years.

When an effort is made to incorporate an existing routine as a component of innovative ro utines, it is helpful if two conditions are satisfied. One is that the routine be reliable-that is, fully under control. The attempt to develop an effective new combination ordi­ narily involves a substantial amount of trial-and-error search, in which obstacles to effective performance are detected, diagnosed, and solved. It is helpful if the familiar elements of the new combina­ tion do not themselves contribute problems, particularly if the problems from that source would complicate the task of detecting and solving the problems arising from the novel elements. The sec­ ond condition is that the new application of the existing routine be as free as possible from the sorts of operational and semantic am­ biguities of scope that we discussed in connection with individual skills. Ideally, the existing routine may require only symbolic repre­ sentation in the design effort for the new combination. For example, the existing routine for shipping the product to wholesalers may be as unambiguously applicable to the new product as it was to the old . In that case, the design effort for the new routine can handle the transportation problem simply by using the phrase “ship to ware­ houses,” and the details of the shipment process need not be exam­ ined. But perhaps the new product is in some way more delicate than the old-more vulnerable to temperature extremes or to vibration. Then ambiguity may arise as to whether the existing shipping rou­ tine will suffice . If there is reason to doubt that it will, the problem of getting the product to the warehouses in good condition becomes in­ terdependent with the rest of the design problem, and the simple symbolic reference to shipment will have to give way to consider­ ation of details. The existing shipping routine may have to be tried out to see how it affects the new product; it may require modifica­ tion, or perhaps the design of the product will have to be altered to make it less delicate.

These two conditions suggest an important qualification to the general notion of an opposition between routinization and innova­ tion. Reliable routines of well-understood scope provide the best components for new combinations. In this sense, success at the in­ novative frontier may depend on the quality of the support from the “civilized” regions of established routine.

3. Heuristics and Strategies as Routines

Our final point concerning the relationship of routine behavior to in­ novation is centered on a simple distinction between organizational activity directed to innovation (or problem-solving more generally) and the results of such activity. The fundamental uncertainty sur­ rounding innovative activity is uncertainty about its results . True, there may be considerable uncertainty, when the activity is initi­ ated, about the details of the activi ty itself-particularly since those details may ultimately be recognized as an approach to some type of success that is not knowable in advance. But there may also be strong patterns of a highly predictable nature in the activity- and to the ex­ tent that this is so it seems reasonable to describe the activity as “routinized.” A particularly clear illustration of the significance of the distinction is the case of systematic sequential search of a well· defined population for an element with attributes that make it the so­ lution to a well-defined problem. When and whether a solution will be found may be quite uncertain, but the search i tself follows a rou­ tine with a simple structure: s elect element, test for desired attri­ butes, terminate with success if attributes are present, select next ele­ ment if they are not.

Routinized arrangements for producing innovations and solutions to problems take a variety of forms,  among which are some very familiar features of the organizational scene. Given a problem, direct a subordinate to look into it- or appoint a committee or a task force, or bring in a consultant with a good reputation. Given a decision to devote 4 percent of $100 million of sales to R&D, it is almost certainly pOSSible to acquire some sort of facility, a research director, and some scientists, and go to work. In broad terms, at least, the art of deploying resources to try to bring about some result or other is not esoteric.  Whether useful results are actually  achieved is  another matter. In fact, results that are more or less  useful are often achieved-and it is an important feature of these problem-solving situations that the superior results that in some sense “could” have been achieved are usually not available as a standard of comparison. The theory of heuristic search provides a helpful framework for thinking about these issues. A heuristic is “any principle or device that contributes to the reduction in the average search to solution” (Newell, Shaw, and Simon, 1962, p. 85). Some heuristics are appli­ cable across very wide ranges of problems-“work backward from the goal”-while others are relevant only in highly specific problem contexts. Devices like directing a subordinate to look into a problem, or appointing a committee, can be viewed as general types of mana­ gerial problem-solving heuristics. But every field of specialized com­ petence contains a wide range of heuristics that are particularly appropriate to that field . The operations researcher will build an op­ timization model of the problem. Th e mechanical engineer will look at the mechanized aspects of the production process, and look for ways to mechanize it further. The chief executive officer whose back­ ground is in finance will bring a different set of heuristics to his job than one whose background is in production . Th e manager who transfers to a new organization will bri ng with him some of the heuris tics that seemed to work in his previous employment.

The broad ideas that shape the most critical high-level decisions of a business enterprise may also be viewed as heuristics – they are principles that are believed to shorten the average search to solution of the problems of survival and profitability. Much discussion of heuristics of this sort has been carried on under the rubric “corporate s trategy. ” Indeed, according to the concept of s trategy that has been <ieveloped by a number of investigators associated with the Harvard Business School, the fundamental heuristic imperative for top man­ agement is: “Develop a strategy.” Other heuristics are involved in the implementation of that basic one-for example, “Assess the company’s strengths and weaknesses in relation to the competition.” A related idea is that the firm should adopt an organizational s truc­ ture appropriate to its strategy. More generally, principles that offer guidance for the selection of organizational structures may be viewed as another class of high-level managerial heuristics.

We propose to assimilate to our concept of routine all of the pat­ terning of organizational activity that the observance of heuristics produces, including the patterning of particular ways of attempting to innovate. To the extent that such patterning persists through time and has implications for profitability and growth, it is part of the genetic mechanism underlying the evolution ary process. But we em­ phasize, once again, that viewing innovative activity as “routine” in this sense does not entail treating its results as predictable.

In many ways our position regarding these matters is consistent with that of Whitehead (1938), who proposed that sometime during the nineteenth century man invented the art of inventing, and is also consistent with the Schumpeter of Capitalism, Socialism, and Democ­ racy (1950), who proposed that sometime during the twentieth cen­ tury the modern corporation “routinized innovation.” Neither Whitehead nor Schumpeter, we think, would deny the role of genius or luck, or argue that systematic differences in innovative compe­ tence do not exist. But their views are quite compatible wi th the proposition that organizations have well- defined routines for the support and direction of their innovative efforts.

Source: Nelson Richard R., Winter Sidney G. (1985), An Evolutionary Theory of Economic Change, Belknap Press: An Imprint of Harvard University Press.

Leave a Reply

Your email address will not be published. Required fields are marked *