Building a Strong Culture

How can we in the software world implement these three values and build a strong culture around them?

Three Leadership Needs

First, we need leadership by example. As a leader, it is part of your job to reinforce and propagate the culture and values of the enterprise, starting with integrity. Our leaders should be held to an even higher standard of integrity than the one to which we hold everyone else. And here I mean all our leaders, not just the folks at the very top. Leadership is part of the company culture, and regardless of where you are in the corporate hierarchy, being a leader means heightened responsibility. The only way to demonstrate integrity as a leader is to act with integrity in everything you do.

It's shocking that the United States Congress has felt the need to pass legislation[11] requiring American CEOs to swear to the veracity of their financial statements. Shouldn't we just expect these statements to be accurate? "Cooking the books" is the moral equivalent of counterfeiting, and Dante in The Inferno put counterfeiters in the bottom of the eighth circle of Hell for good reason. His conception of Hell consisted of nine concentric circles, so being at the bottom of the eighth was almost as low as you could go.

[11] The Sarbanes-Oxley Act of 2002.

Second, to develop a customer-focused culture, our leaders need to spend time with customers to gain an acute sense of what the customers need. Only then can they translate that into direction for the rest of the organization.

Finally, we need leadership in the area of pay for performance. Managers at software companies need to be able to objectively evaluate performance and compensate people for their contribution to the organizationand nothing else.[12]

[12] See Chapter 16, "Compensation."


In addition to meeting these three needs, we need continuity. Cultural propagation of values is meaningful only if it survives all the varieties of change that occur in the business world. Let's have a look at some of the phenomena that threaten continuity in software companies.

That Old Devil, Growth

Inevitably, successful companies grow, however small they are at first. Their culture needs to be nurtured so that these organizations do not abandon the very values and traditions[13] that made them successful in the first place. But it is not easy to do this right.

[13] A tradition is a cultural artifact that becomes ingrained, so that people observe the tradition without even remembering the underlying value that created it.

If your company grows slowly enough, there is hope. You can recruit very selectively, working hard to find people who are "good cultural fits" for the existing organization. Rapid growth, on the other hand, is a double whammy. In the scramble to fill open positions,[14] it is easy to lose discipline and get sloppy. On the margin, we hire people we probably wouldn't if we weren't in such a hurry. Then, to make matters worse, we don't explain to them the cultural milieu into which they are about to be thrust because we are too busy hiring the next batch of folks.

[14] One of my least-favorite expressions in the business world.

Notice that we are not necessarily recruiting people with the wrong values. No, what typically happens is an overall weakening of the culture as the organization expands. For example, occasionally forgetting to put the customer first may not draw sharp criticism any more. Over time, Gresham's Law takes over: The bad money drives the good money out of circulation.[15] Those living by the old values become the minority, viewed at first with curiosity, then with derision, and finally ignored.

[15] This needs some explanation. Sir Thomas Gresham, advisor to Queen Elizabeth I, noted that when a new currency is introduced to replace an older, debased one, people tend to hoard the new currency while trying to spend as much as possible of the old. This is the origin of the expression, "Bad money drives out good." In a business context, though, we can borrow this old saw to describe what happens when a small company with a strong culture undergoes the transition to becoming a large company with a weak culture. The newer, weaker, more diffuse and less desirable culture is shared by many more people so that, over time, the older, stronger, more valuable culture is marginalized to a smaller and smaller number of adherents, eventually dying out. The "tyranny of numbers," unfortunately, wins out.

What replaces the strong culture is often a bureaucratic system designed to check on and enforce a wide range of policies and procedures that never quite capture the previous spirit of "do the right thing." In the absence of cultural strength, organizations fall back on meticulous attention to detail, as if we could recapture the spirit of the law by increasing the number of its letters. Leaders assume that everyone else will make the wrong default decisions, so they impose lots of rules to guide and correct them. It's a sure sign that bureaucracy has set in when the Ten Commandments are replaced by a 247-page handbook that sits on a shelf gathering dust.

Mergers and Acquisitions

After growth, the second most destructive force for a strong culture is a merger or acquisition. The software industry periodically goes through cycles of consolidation; many companies, Rational Software among them, grow "organically" for many years and then use mergers and acquisitions as a way of continuing or accelerating growth. Unfortunately, software companies seem to be much worse at this than other industries, perhaps because software companies have such a wide spectrum of cultures and values. The largest single reason for failure when two software companies combine is cultural incompatibility. Even if the two cultures are similar, merging them can be difficult for a vast variety of technical reasons. Plus, if the two companies are located some distance from one another, there is insularity because of the separation. Whatever the root cause, in the face of fundamental incompatibility most software mergers fail, plain and simple.[16]

[16] Wall Street confirms this almost every time a software merger is announced. The typical response is a decline in the share price of both companies prior to the accomplishment of the merger. Only after the merger has proven its value does the share price of the combined company recover. While this phenomenon is also observed in other mergers (even non-technical ones), it seems to be more severe in the case of software companies.

There is a solution. When our pioneer great-grandmothers crossed the prairie in their Conestoga wagons, they always carried a lump of "starter dough" so they could bake the same kind of bread in their new home that they did in their old location. The starter dough contained a yeast culture[17] that could spawn new loaves; every time they baked, they saved a bit of that dough to mix with new dough later.

[17] According to the dictionary, the word culture is used for both "the customs, institutions, and achievements of a particular nation, people, or group" and "the cultivation of plants, breeding of animals, or production of cells or tissues." I leave it to you to decide whether this linguistic overloading is a mere coincidence, or whether it belies something deeper.

Similarly, by transferring one or more people, preferably senior people, to the new location, software companies can provide "starter dough" to transmit the company culture and values to the newly merged or acquired company. Any merger or acquisition that forgets to transmit cultural messages early and often is a disaster waiting to happen.[18]

[18] In a stunning bit of irony, as this was being written in late 2002, Rational Software was itself being acquired by IBM. The acquisition was announced almost simultaneously with the original publication. So far, it seems to have gone very, very well. I'm keeping my fingers crossed.

The Single Big Customer or Partner

Another interesting threat comes from having a single large customer or partner with a very strong culture. In this case, both subtle and not-so-subtle influences can permeate your organization. For example, the customer's or partner's style of reporting resultsnature, frequency, and so onmay be imposed upon certain projects; gradually, it may spread through the ranks, so that all internal divisions start reporting in the same way.

Sometimes this can be good; we can always learn from others. On the other hand, we do need to be careful that the customer or partner shares our values and that what morphs is the cultural manifestation of those shared values, not the values themselves.

Finally, it is an unfortunate fact of life that partnerships in the software business world are complicated. Today's partner may be tomorrow's competitor, and vice versa. Figuring out the implications of that one is an exercise left to the reader.

New Efforts

Yet another threat to a strong culture is the startup of a new effort, such as a new product line or a facility in an overseas country. Here the "starter dough" principle is crucial. I have personally witnessed a new organization built from the ground up halfway around the world that faithfully reproduced the "mother culture," simply because it started with one person who thoroughly understood the original recipe. He hired and trained every single addition. Ten years later, through slow but steady growth, this organization was one of the strongest in the company, both culturally and in terms of productivity, even though they were geographically farthest of all from the home office.

Be especially wary of efforts to start new groups whose express purpose is to launch a radical cultural change; these are almost always doomed to failure. Here's why: If the change is that important to make, hitch up your pants and do the hard work to make it across the entire organization. If you don't, the new organization will always be on the outside, resented by the rest of the company. If they are successful, they will be resented even more. As they grow in power and influence, civil war will loom, and the only solution may be to spin them off. In so doing, all the benefit to the original organization will be lost. On the other hand, if the experiment fails, it will likely be because the splinter group felt (or was) orphaned.

The Software Development Edge(c) Essays on Managing Successful Projects
The Software Development Edge(c) Essays on Managing Successful Projects
Year: 2006
Pages: 269 © 2008-2017.
If you may any questions please contact us: