Best Practices

I have seen a lot of confusion around the term, “best Practices” lately. In testing, this confusion usually results in many hours of unnecessary work, which also results in project delays.

When project groups are asking for best practices in a new process, what they are really asking for is how other universities are doing the same thing. We are finding that they are all different, so there goes the concept of a best practice. We still need to find the best way for us, but it is NOT a best practice. If it were, we would all be doing it the same way. It may be the best practice for us, though, so I prefer to call it a “good practice in context.”

This is especially true in testing. If you are familiar with the “What is QA?” page on this site, you know that our approach to testing is Context-Driven. Which simply means that there is no one best way to test anything. The testing strategy is fully dependent on the context of the project. After all, if there really were one best way to test, there wouldn’t be so many other companies looking for their own best practices. 

“There are good practices in context, but there are no best practices.”

This month’s post is from a Better Software Magazine article published by  Michael Bolton does an excellent job dispelling the myth of best practices. Read on to see why he believes there’s no such thing as a best practice.

Comparatively Speaking

by Michael Bolton

While I was waiting in line at the bank last week, I overheard a conversation between one of the tellers and a customer.

“What’s the best account for me to open?”

“Well, it depends on what you want.”

“I want to open the best account you have. Which one should I open?”

“Sir, it depends on what your goals are. To manage your cash flow, we have checking accounts that give you detailed monthly statements and low charges for checks and automatic withdrawals. For a high interest rate on your deposits, we can offer you savings accounts. Credit card accounts also can be useful; they have differing interest rates and differing payment terms, depending on your needs. Or, if you’re interested in buying a house, we offer fixed-rate or floating-rate mortgages. The choice is yours.”

The question about the best bank account sounds ridiculous, doesn’t it? Yet, many customers ask me about “industry best practices” as though there is a single software industry or practices that are “best” in all contexts.

Jerry Weinberg suggests that when we hear comparative words such as “fast,” “expensive,” or “good” we should immediately ask, “compared to what?” So when we hear “best,” we should ask:

“Best” for whom? Each existing practice provides benefit to some part of the organization. Who’s asking for best practices? Who will get the benefits? Who will be required to adapt?

“Best” for achieving what goals? If our organization is achieving some goal today, aren’t we already using a best practice? What goals might require different practices? How might those practices have an impact on the goals we’re achieving now?

“Best” according to whom? Do the people who are proffering the “best practice” understand our business? Do they even care about our business? Are they selling us something? Do they have an axe to grind?

“Best” as measured by what yardstick? Do we measure things in the same way as those recommending the practice? How will we know when the best practice is providing benefits? Are we willing to reject the practice if we don’t see benefits?

“Best” in what context? Who are the people already using the practice? Are they really using it? Is their business really the same as ours? If we want to be the same as these people, why aren’t we already that way?

“Best” at what cost? Would what we need to apply—time, money, effort, or staff—and what we may have to give up—flexibility, time to market, or, in the case of some Agile processes, accountability—be worth it, or might it undermine the reasons that we’ve been successful so far?

All things—cars, movies, bank accounts, processes—have attributes that someone values. Someone may disagree, for perfectly valid reasons, with anyone who identifies anything as “best.” Practices should be our servants, helping us get the job done, and not our masters, dictating what we must do. On their own, our practices don’t produce anything; our people do. Hmmm … isn’t it odd that no one has ever asked me about “best people” methodologies?

File: Download PDF

I sincerely hope that this post helps to clear up some of the confusion around the term “best practices.” Please post any questions or comments below. Thank you!