Agile Decision Making - When Wrong is Good
Byron Jung
Project Management and Agile Leadership for Data and Software Engineering (PMP, Certified Scrum Professional, Certified Facilitator)
Without a doubt, all successful projects rely on good decisions. If you ask, most people will equate good decisions with right decisions.
Good = Right?
I would also suggest that good decisions depend on how quickly they are made. Undoubtedly the best decisions are those that are right and made quickly. We can probably agree that the worst decisions are the ones that take a long time to make and turn out to be wrong
What about the decisions that are fast but wrong?
Most of us have heard some variation of the advice to “fail fast and often”, which recognizes that right decisions are hard to come by. Failing fast allows time for experimentation to ultimately arrive at a right decision. Agile is designed to tolerate mistakes. Chances are good that if your organization subscribes to Agile, trial and error by the software development team is cheap and mistakes aren't fatal.
For Agile organizations I’d argue that a wrong decision made quickly is a good thing because it allows additional time for trial and error.
On the other hand...
I would argue that a bad decision for Agile organizations is one that’s right but took a long time to make.
There are 2 reasons why right, but lengthy, decisions are bad.
- Trial and error would likely have produced the right decision much faster than lengthy deliberation (think of learning to bike by riding versus being on the sidelines studying other people riding)
- Another, possibly more insidious, reason is that right but lengthy decisions tend to condition an organization into thinking that all correct decisions are a function of spending a lot of time and that wrong decisions are results of not spending enough time. Invariably, when wrong decisions are made - despite much analysis - these organizations will take an equally long time to make the right decision. So in a given period of time they will make fewer right decisions than organizations that can cycle through decisions faster.
Food for thought.
If you belong to a self-professed Agile organization that finds itself spending a lot of time doing analysis with the goal of making right decisions, you should be aware that bad decisions are likely the output. You will be overtaken by organizations that can make more right decisions in the same period of time.
If a decision for your Agile organization is simply too risky for trial and error at a particular point in time, then consider avoiding a decision altogether so that you don’t prematurely or arbitrarily limit your options too early. Being a cousin to Lean manufacturing philosophy, Agile is inherently tolerant of saving decisions for later (e.g. being able to add new requirements to the backlog while implementation is already underway). Instead, consider how to focus your energy on other decisions that can be made to propel your organization forward.
In an upcoming post, I will share my thoughts on what constitutes a fast decision versus a slow decision. Your comments and feedback are welcome.
Digital Technology Executive
9 年Great article, thanks for sharing your thoughts! I feel like you are missing a dimension in your graph which is "Reversibility". In my personal experience I have found that it's one of the most important factors to consider when you are making an important decision as an individual, team or organization. Majority of decisions we make are reversible and in this category of decisions speed always wins as you have pointed out in your article. One thing to remember is that a vast number of decisions product teams make are made under conditions of extreme uncertainty and should be treated as experiments validating a certain hypothesis. The hardest decisions to make are: which hypothesis is more important to validate next, how do we measure the outcome of the experiment and what needs to get undone if we are not happy with the results? However, there are some decisions that are difficult or impossible to reverse and it's OK to invest more time into analysis and potentially involve more people in the process that you would typically have for reversible decisions: launching or sunsetting a product, acquiring a company, hiring or firing a team member, signing a long term lease agreement for an office space or investing into a new tech stack. You still want timebox those but allow enough time for analysis depending on the scale of the decision. Fast decision-making is really an imperative for those companies that want to outlearn and outperform their competition. It's important to turn it into an organizational habit at every level so it becomes a default modus operandi overtime.