words from a tech maven

This is a presentation prepared for the August 2010 Austin Product Camp event. Here is the description:

Engineers in product development typically find many things to grumble about. Besides the usual “too many features, not enough time” they may feel underappreciated as people and put off by inappropriate styles of management. In large part this is because engineers, like other creative workers, are “right-brain” oriented, while management is mainly a “left-brain” activity.

An experienced engineer who has “been there” suggests that excessive left-brain stimuli (meetings, demands, etc.) can interfere with creative work by blocking right-brain concentration. If managers could better appreciate this insight and adopt more right-brain-friendly management practices, they would be much more effective as leaders building product teams willing to “go the extra mile” to achieve a successful result.

The PowerPoint file (94 KB) can be found here.

(This post is based on an online comment to an article (posted May 14, 2010) by Prof. Uwe Reinhardt of Princeton University in the “Economix” blog of The New York Times, titled Companies: What Are They Good For? )

The classical economic view of a business, based ostensibly on the writings of Adam Smith, is as the property of its owner, who benefits his customers, and thus all of society, by running the business so as to maximize his own wealth. Today this model has been extended to treat public companies as the “property” of their stockholders, who alone have a say in the management policies of the company by virtue of their exclusive property interest. Management’s sole obligation is therefore to satisfy the stockholders’ demand for profits (echoes of Cornelius Vanderbilt’s “The public be damned! I’m working for my stockholders!”).

In contrast to this, the “stakeholder” model considers the interests of all the parties involved in the operation of the company, seeing it as a coalition among:

If I were to include a diagram here, it would show these as the arms of a cross, with management in the center, coordinating operations and determining the share-out of profits. The company is of course itself a customer to its suppliers, and management are also themselves employees. Note, however, that all revenue comes from customers, and all value-adding input comes from employees in the form of labor (including creative and intellectual work); therefore it is not in the company’s economic interest to stiff either one for the sake of supposed economy. Society, government, etc. are of course also involved as interested parties but not as direct participants in the operations of the company.

Many of the on-line comments to Prof. Reinhardt’s article rightfully point out that the “ownership” of modern corporations is actually very diffuse, or that the say of shareholders is very limited; also that there is a general opinion that the mission of a company is to serve the public no less than the stockholders. So it would seem that the rationale for the classical model, at least among non-economists, is not very well supported.

Prof. Reinhardt cites an article by Klaus Schwab, “founder and executive chairman of the renowned Geneva-based World Economic Forum,” in the Wall Street Journal of January 14, 2010 as defending the stakeholder model; indeed, Schwab in the article claims to have originated the idea and first presented it some 40 years earlier at the World Economic Forum in Davos. Something very troubling, however, is the tone of the on-line comments to Schwab’s article: the main issue raised here (in contrast to the comments to Prof. Reinhardt’s article in The New York Times) is the objection that the stakeholder model gives no consideration to property rights. Customers, suppliers and employees supposedly should have no say because they have no property stake in the company. Judging from the language used, it would seem that there is political baggage involved: defense of private property is apparently seen by many as a bulwark against communism, toward which communitarianism (e.g., as reflected in the stakeholder model) is just the first step down a slippery slope. I see this argument as equivalent to saying that the right to vote should be conditional on the ownership of property; or, taking it to an extreme, that if someone is beating me with a whip, I have no right to protest unless I own the whip – or if someone shoots me, there is no crime committed as long as the shooter owns the gun.

I would argue (although I am not an economist) that we badly need a new analysis, and possibly redefinition, of the nature of property and ownership appropriate to our current conceptions of value, utility and infringement of the personal rights of others. It is clear to me that, for example, with “intellectual property” many of the usual behaviors toward property are not appropriate (see this post below). What do we mean today by “property” and “ownership” in view, for example, of the fact that “possession” may not be exclusive, or that the uses of property are circumscribed by various legal and ethical provisions? An updated theory of property and ownership would address these issues.

To be an innovator, learn to think like a designer.

Design is not a frill, or just a matter of aesthetics. You can design, for example, a chair so as to be aesthetically pleasing, in the style of Frank Lloyd Wright, or Eero Saarinen, or Le Corbusier, or whomever, but it may not be comfortable to sit in; you will have produced not a chair, but a sculpture. On the other hand, if you start by making the chair comfortable, chances are it will also meet the aesthetic standard.

I would make the following statement: Good design is a matter of maximizing the suitability of an object to its intended purpose, while at the same time satisfying certain constraints as to cost, choice of materials, fabrication processes, etc. An object thus designed will also likely be pleasing in appearance.

The above applies to non-physical objects as well: for example, the user interface of a Web site or a software application. In this case “maximizing the suitability” amounts to minimizing the “cost” in mental effort required to navigate and make selections so as to accomplish an action. Today this effort can even be quantified by actual measurement of brain activity, e.g. using MRI scanning techniques.

Besides the user interface, there is also a level of “meta-design” representing the structure of the process being implemented. A Web form basically implements a decision tree; how many branches are necessary, bearing in mind that each branch may represent another mouse-click?

Learning to think like a designer means learning to see everything in terms of its suitability. The chair you are sitting in: how comfortable is it? The Web form you are entering information into: how hard do you have to work to navigate it? Your kitchen cooktop, or washing machine, or the smartphone you just bought: are the layout and controls intuitive, or do you need a user’s manual just to see how to turn it on?

When you start looking at things this way, you start to notice and discriminate instances of good vs. poor design. And when you realize that every instance of poor design is an opportunity to rethink something to make a better product, or replace it with something completely new, you have the chance to become an innovator.

Abraham Lincoln is said to have posed the rhetorical question: “If we call a tail a leg, how many legs does a pig have?” Five? “No, four. Because calling a tail a leg doesn’t make it a leg.”

Consider what happens when we try to deal with new technological concepts in terms of pre-existing legal, business or other non-technical categories. One of the best instances of a concept which is “new” – in the sense of not directly corresponding to anything that existed previously – is that of software. To get an idea of the confusion that results, look at the legal debate as to whether software is “speech” (hence protected by the First Amendment) or “action” (hence not protected). Any technically-minded person who deals with software on a daily basis knows that it is neither; nor is it “art” (although aesthetic judgment plays a role in its creation) or “literature” for that matter.

A further example of categorical confusion arises when the concept of “intellectual property” is applied to the information resources (including software) generated in the course of designing and producing a marketable product. The point here is that just because something is called “property” does not mean that all of the conventional behaviors associated with the handling of property are appropriate to it; to paraphrase Honest Abe, just calling it “property” doesn’t make it that.

The essential attribute of “property” is exclusive possession which in itself has value. “Property” is something for which a price can be asked and received in exchange for the transfer of exclusive possession. This in turn implies the possibility of theft, as a way of getting the benefit without paying the price. To prevent theft, “property” must be guarded and access to it restricted. This is part of the “conventional behaviors” we associate with “property” as we normally conceive of it.

What determines the “value” of an item of property? Is value intrinsic? People thought so in the classical (Greek/Roman) period; the value of an object usually depended on what it was made of. A thing made of gold was considered more valuable than one made of, say, copper. In contrast, today our concept of value is determined by utility: what can a thing be used for? what is the cost of not having it when we need it? (A third, perhaps Marxist, concept of the value of a thing is the cost of the labor and materials that went into producing it. To illustrate, consider the “value” of the plutonium recovered through the dismantling of nuclear weapons in the wake of the Cold War. For the utility-minded Pentagon, the stuff is a less-than-worthless nuisance: practically useless, highly noxious, hard to dispose of, and a security risk besides. The Russians, on the other hand, see their plutonium as a “national treasure” in view of the cost and effort involved in producing it.)

For our purposes, we can define “value” from a seller’s viewpoint, as the price which maximizes the ultimate economic benefit to the seller. If the price is too high, there will be no buyers and the economic benefit is zero. On the other hand, if the price is zero or even negative (paying people to take something), there can still be an ultimate economic benefit, as we shall soon see.

So what does all this have to do with “intellectual property” and software? The question I am raising is whether software should always be treated as valuable “property” to be sold (licensed) for a price, and therefore requiring protection from theft. The same could also apply to other information resources (e.g., interface specifications) generated in the design process. In regard specifically to software, the question is all the more of interest today due to the existence of the “open software” movement. The proponents of “open software” claim that, at the least, users of software should have access to the source code, to be able to modify it so as to customize features and to fix any bugs they may find, and that they should be free to make their modifications available for the benefit of other users. This requires that source code be freely accessible and that users not be restricted by proprietary considerations. A stronger, more ideological position (associated mainly with Richard M. Stallman) is that software itself should be free on principle and accessible to all at no cost.

From the “property” viewpoint, “free software” amounts to a giveaway of that which could be sold at a price. On the surface, this makes no sense in business terms; at the least, if users can modify the code and make their changes available to others, it appears to mean giving up control. Against this, the “open software” position is that with many eyes on the code, bug fixes and user-driven refinements will be made in less time, resulting ultimately in stronger and more useful software (the success and reliability of the Linux operating system are usually mentioned here as a supporting example).

It would seem that there is no middle ground: either (a) treat software as a revenue-generating protected asset, to be sold (i.e., licensed) for a price on condition that it not be modified, reverse-engineered, etc.; or (b) make the source code public, and abandon any hope of generating revenue from the software itself (as opposed to sale of support services).

There is actually a third possibility, to “leverage” the open-software process so as to enhance the value of proprietary products. To illustrate how this might work, consider this example: A manufacturer of hardware controllers (e.g., sound cards for PC’s) might publish its specifications for interfacing with software drivers, and encourage users to develop (on an open-source basis) actual drivers for different operating systems. The driver code would remain open-source, freely available; there need be no licensing conflict with the manufacturer of the hardware. It would even be possible to support the open-source development work financially, through grants to teams of developers, provision of equipment for their work, etc., on terms which generate no legal “strings” (e.g., it would be clearly stated that no implied warranty is created through such support).

From a naive point of view, it would seem that the controller manufacturer in this example, in freely publishing its specifications, is giving away a part of its “property” and thereby losing possible revenue. However, the overall effect would likely be to make the company’s hardware more attractive to the market, as a result of the wider variety of drivers available. This is precisely the “leverage” effect I am proposing.

Another argument that might be raised against giving away such normally “proprietary” information, is that by doing so the hardware manufacturer (in the above example) is making it easier for others to create a competing product for the same market. But the likelihood of this happening should be assessed realistically: what is being given away is only a small part of the complete design, and even if a competitor were able to derive benefit from it, a long lead time would still be needed to design the competing product and get it into production. Thus the competitive advantage being lost would be small, and would likely be more than compensated by the benefit from the “leverage” effect of open-sourcing the driver development.

The effect would be the same if the proprietary product were a software package, e.g. a content management system or CMS for building Web sites, and the object were to open-source the development of a variety of add-on and plug-in modules for displaying material in different formats. The company offering the proprietary CMS would “give away” a minor portion of its “property” – the plug-in interface specification, in this case – and benefit many times over from the resulting wide range of plug-ins which enhance the market appeal of the CMS itself.

In this situation it is interesting to consider: what is the effective seller-side “value” of the “property” which is given away? The strange thing is that, contrary to the usual “property” model, giving it away may in the end create greater economic benefit than trying to generate revenue from it through licensing. The number of hobbyists, for example, who would take it for free and make productive use of it would likely be more than the number prepared to pay the licensing fee. So the effective “value” of the interface specification as “property” in the conventional sense might well be negative! The CMS company would be crazy to not give it away. This is an example of how forcing a new concept into a conventional category – in this case, thinking of an intellectual resource, here a design specification, as “property” – ends up being economically counterproductive.

The point I am making should be clear by now: The “open software” movement is a source of potential benefit to sellers of proprietary products, through the “leveraging” process as I have proposed it. Companies should not be afraid of letting go of the intellectual capital they generate in the process of creating new products, but should instead treat it as “seed” to foster parallel development, in the open-source arena, of add-ons and other utilities which enhance the market attractiveness of their prime products, but which are perhaps only secondary to the companies’ own product development strategies, or would divert resources that the companies would prefer to dedicate to prime products. By establishing a cooperative relationship of this kind with its user community, a proprietary-product vendor ultimately benefits more than by taking a strict lock-the-barn-door attitude toward its intellectual assets.