Agile is not a tool

Many times—and it would be hard to try to count at this point—I’ve sat with a team to discuss how we’re working together, and someone asks the question:

What is stand-up for, anyway?

What makes this question so hard to answer? Why do we spend so much time doing something that we aren’t even sure about? Today I bring you those answers and more as we unpack everything that’s wrong with agile software development.

• • •

In the 1970s, an enterprising cardiologist read some research that was carried out at DuPont during World War II and concluded that restricting carbohydrates could be the “key” to weight loss. He wrote a couple of books about it and not much happened.

In the 1990s, he formed a company called “Complementary Formulations” and started selling supplements and snacks with a low-carb theme. People started to notice and business picked up. In 1989 the founder changed the name of the company to his own: Atkins Nutritionals.

By the early aughts, about one in eleven North American adults were reportedly on the Atkins diet. If you were around back then, you’ll remember the persistent buzz around this “diet revolution.” It worked, too. At least, at first. There’s some debate about the underlying mechanisms, but practically speaking people who go on Atkins immediately start losing weight.

Unfortunately, this low-carb, high-fat approach isn’t always great in the long-term. Just ask Jody Gorran, who sued the Atkins estate in 2004 after he had a heart attack caused by high cholesterol. Gorran had been on the Atkins diet for two years and his LDL got so high he had to have a stent installed.

The mistake Jody Gorran made was to assume that because a large and sophisticated enterprise existed around the Atkins diet, and because at least one in ten of his friends were doing it, it was both safe and effective.

I’m sure he felt differently after waking up from his angioplasty.

The moral of this story, if there is one, is that the marketing success of any solution confirms only one thing: people have the problem it is trying to solve. That’s it. Americans are in a constant state of frenzy about losing weight and they’ll try anything. The success of the Atkins diet (and South Beach, and Weight Watchers, and Jenny Craig, and so on) is a confirmation of that fact alone. Solutions can enjoy wild success even when they don’t work.

Dr. Atkins’ interpretation of previous papers was reasonable, and there is some validity and efficacy to the Atkins diet, but it only got dangerous when he commercialized it. Transforming nuanced scientific research into a product gave consumers the impression that this was basically a no-downsides, turn-key solution to their diet woes.

• • •

But what does all of that have to do with software development?

“Agile” software development is almost mythological at this point. Many of us move through our days performing the motions that we have been told are “the agile way” and that all of this is “so much better than waterfall” while we don’t even understand where it came from or why it matters.

If you’re doing planning meetings, stand-ups, and retrospectives across approximately two-week “iterations,” what you’re doing is called Scrum, and Scrum, my good friends, is a brand.

Ken Schwaber and Martin Fowler were among the folks who wrote the Manifesto for Agile Software Development and kicked off a deliberate push toward the agile way of working. Schwaber took it one step further and formed the Scrum Alliance and started peddling his specific implementation of agile principles to companies.

Just like the Atkins diet, Scrum is built on foundational ideas that are reasonable and useful. The Agile Manifesto is some of the best advice we have for navigating the totally unpredictable cycle of software development. But Scrum itself is basically a shrink-wrapped version of agile created to make the thing “sellable” so Schwaber could take it into companies and get paid.

Scrum is a wildly successful product. There are over 300 Scrum-adjacent trademarks filed with the US Patent Office and brand recognition for the term “Scrum” in software circles is nearly 100%. Only the most iconic brands reach that level of awareness. Scrum is a triumph.

Scrum mostly works, too. I mean, it’s better than anything we tried before, and I’ve worked with teams doing some flavor of Scrum for about 20 years and we got shit done, so what’s the problem?

It’s people. People are the problem.

• • •

No two people are alike, no two teams are alike, no two companies are alike. It is for that exact reason that the very first value in the Agile Manifesto is “[we have come to value] individuals and interactions over processes and tools.”

Scrum is a process, and it drags along with it a metric shit-ton of tools like Jira, Trello, Monday, Asana, Pivotal Tracker, and on and on. Just like the Atkins diet, the Scrum industrial complex needs you to believe that the only thing standing between your team and shipping tons of great code is a Jira site license.

That is demonstrably false, but the secret to getting rich in a gold rush is to be the one selling the shovels, and with global software spending exceeding one trillion dollars a year, everyone has a shovel to sell you.

Indulge me as I stretch this metaphor to its breaking point. You can’t dig for gold without a shovel. Needing a shovel and buying a shovel is neither wrong nor bad. Most of these shovels work, too; Jira works, Trello works, Asana works. You can configure them to be terrible, but that’s your fault. Just because the shovel has a certain shape of handle or comes with a certain set of instructions doesn’t mean those things work best for you.

There is such a variation in teams and projects that accepting “the default” of what Scrum prescribes is almost guaranteed to be less than optimal. Scrum is successful because it’s 1,000 times better than the non-agile approaches that came before it, but that doesn’t mean it’s the best for you and your team.

• • •

So what are stand-ups for, anyway?

Stand-ups are whatever you want them to be. That’s either a disappointing answer or an inspiring one, and how you feel about it right now might tell you a lot about how engaged you are with your development process.

The most basic underlying philosophy of the agile way of working is creating high-fidelity feedback loops over relatively short experimental periods and deliberately collecting and using that feedback to change the way things are done. That’s it, that’s the entirety of agile in one sentence.

The enterprise value of agile is to quickly and continuously converge the software project toward the greatest end-user value by shipping iterations quickly, gathering direct feedback from the end-users, and reshaping the roadmap based on all available information.

The team value of agile is to do the same thing for the way in which you do it. With each iteration shipped, you learn which parts are working for the end-user and which parts aren’t. You also learn which parts of how you built it worked for you, and which parts didn’t.

Scrum is a perfectly valid starting point for that process, but the secret to success is to think of it that way: as a starting point. If you aren’t engaging your team in conversation about how things are going, what went well, what didn’t go so well, and what could be improved, you’re just enjoying the modest weight loss of the Atkins diet and waiting for the heart attack.

Questions for you

  1. How do you engage your team in deliberate reflection on its process?

  2. What questions do you ask that help you understand how things are going?

  3. What is one assumption baked into your tools or development approach that is worth examining?

Lead image by Midjourney AI

Comments