« Demystifying ITIL | Main | Culture as Infrastructure for Strategy »

September 3, 2005

Optimization - the Effectiveness of Efficiency

optimize

v 1: make optimal; get the most out of; use best; "optimize your resources" [syn: optimise] 2: modify to achieve maximum efficiency in storage capacity or time or cost; "optimize a computer program" [syn: optimise] 3: act as an optimist and take a sunny view of the world [syn: optimise] -- Source: WordNet ® 2.0, © 2003 Princeton University

"Doing More With Less" actually has some competition these days (!), in the call for Optimization.

I toyed with the idea of demonstrating this by citing the number of search engine hits for "optimization", but of course the results of the search would be highly dependent on how many sites have been optimized for search engines -- probably the reason (or at least, a great reason) why the idea of doing optimization has become so widely propagated.

Leaving out the number of site hits, it still seems that almost anything can be "optimized", if you can believe everything you read. The problem is that the point of it usually appears to be to maximize something - so what's the big deal about using the word "optimize" ?

I.

Getting the most out of something really is the main idea of optimization. But while we naturally crave the "most" part of the offer, the lion's share of the practical attention to optimization goes to the "getting" part.

To the point and in summary, in optimization the usual assumption is that something we own or use has a less-than-maximum effectiveness because of some underlying inefficiency.

It helps to know that the final point of optimization is always really about "effectiveness" -- it means that the starting point of optimization efforts is always a determination of what effect the owned or used item is supposed to have. After all, as they say, if you don't know where you're going, then it doesn't matter how you get there...

Then, looking into the dependencies and causes for generating that effect, we'll see opportunities to solve problems, make enhancements, or innovate -- in order to increase the item's effectiveness that we need.

But that is why a problem arises with the popular conception that optimization is synonymous with "efficiency". There might not be a problem if the efficiency issue was always driven by an explicit objective defined by effectiveness. But, unfortunately, by analogy, while a square is always a rectangle a rectangle is not always a square. That is, optimization can be an outcome of efficiency, but "efficiency" does not always produce optimization. The problem arises in the ideas about efficiency and especially in myopia about optimization needing efficiency.

For example, getting further into that point, how would we understand activities like corrections, enhancements and innovations to be "efficiency" techniques?

The answer of course must be that "it depends on what is being changed." This realization breaks the myopia and helps solve the problem.

II.

If being effective means "having an intended or expected effect" while being efficient means "being effective without wasting time or effort or expense" then we know we're looking at whether the underlying mechanism of the intended effect is functioning in a way that includes unnecessary waste.

The challenge is to define the waste, confirm that it is unnecessary to the intended effect, and validate that the technique for its removal is not just going to replace one kind of problem with another kind that is equally undesirable.

But let's get a gut-level grip on this matter of what to change. Suppose that you are on a solo hiking trip, not anyplace too obscure or remote, but you get a bit lost or significantly delayed. You pull out your cell phone, but the transmission connection in the area is very marginal and very unreliable. On the one hand, during this attempted use of the phone, all parts of the communication system might be working exactly as intended, perfectly. We generally know that the parts are not malfunctioning. There may be no "efficiencies" to add to the components. But on the other hand, the combination of these components -- that is, the underlying mechanism for connection -- is not up to the task of producing the intended effect. In what sense(s) is the combination "inefficient"? What is being wasted?

This points up some fundamental considerations of "optimization"...

For one thing, when effectiveness is lacking, a major possibility is that the design of the underlying mechanism might be inappropriate, not inefficient. We wouldn't normally use a shopping cart to move a baby grand piano. Bulking up the shopping cart would be solving the wrong problem.

However, inappropriateness can be expressed as "an inefficient relationship" because, due to the intended effect, the tasks that are imposed on the underlying mechanism have performance requirements attached to them that the mechanism only inconsistently meets. These inconsistent support results mean that the opportunity is being wasted compared to what would be obtained from consistent support.

For another thing, it often goes overlooked that the situations we care about optimizing are ones that extend over time -- typically because some kind of action is recurring in order to produce what we want. Over a stretch of time, the recurring action may encounter significant differences in the surrounding circumstances from one point to another. The challenge is for the action to succeed at each point despite the differences in conditions. In fact, it's reasonable to expect that if the action itself doesn't change then its level of success may vary significantly from one point to the next.

If the conditions that are expected during the time period include significant variation, then getting the most out of the action must first be measured not on a point by point basis but on an overall basis for the period. A major characteristic of an optimized mechanism is that because it can make intelligent trade-offs it can persist over time within tolerances: it won't exhaust itself at every single moment if the price of doing that is inability to handle the long run. Resilience and agility is actually a hallmark of a mechanism geared for effectiveness.

III.

The focus on opportunity waste as opposed to production waste is what basically distinguishes optimization projects from mere efficiency projects. Eliminating waste must first be understood at the level of the opportunity, because that understanding will indicate what aspect (if any) of production should be changed and (with the benefit of relevant knowledge) how it should be changed.

At the underlying production level of waste, we deal essentially with the problem of the production mechanism's suitability to task -- both at the overall mechanism level and at the component level. Changes to the mechanism can be technical, logical, or whatever -- but the reason we deal with it is because of the appropriateness of the mechanism to the intended effect. If we live on a cowboy ranch and we have only a minivan, regardless of the condition of the minivan we're not so much concerned with changing the van as we are with getting a jeep.

But once we have a jeep, the issue becomes, "is the jeep good enough?"

While optimization is inextricably tied to issues of efficiency, the "elephant in the room" in discussions of optimization is "quality"...

Raising the effectiveness of a resource, within the tolerances of its assigned circumstances and intent, is actually about Quality. If efficiency is a path to higher quality, then efficiency should be leveraged but with a priority that also recognizes other paths.

But when we hit the limits of efficiency, if we also hit related but unintended consequences -- such as brittle processes, siloed operations, funding shortages, and rigidity against future change -- the net is not more effectiveness. We'll have wound up with something that is very very good at doing something we don't need or causing something that we don't want.

IV.

Optimization efforts should never be considered strictly in the context of efficiency -- because the necessary outcome of changes is always fitness-to-purpose, or quality, and only sometimes fit-to-capacity. We only ever put five players on the basketball court at one time -- but the key to winning (being effective) with those five is not the player-by-player tuneup but instead the organization of all five players in the tasks of the game.

The essence of optimization calls for a broader perspective that most highly prioritizes relevance to objectives -- followed by a range of options for changing the design of the support including corrections, enhancements and innovations. Efficiency, then, should be understood in terms of how to best execute the changes that improve the design.

Posted by Malcolm Ryder at September 3, 2005 6:05 AM

Trackback Pings

TrackBack URL for this entry:
http://www.malcolmryder.com/cgi-bin/mt-tb.cgi/124

Comments

Post a comment