Big Consulting Agile
I've now heard this story from multiple independent sources, working at completely different companies:
Leadership brings in a big consulting company to "help with efficiency"
The consultancy introduces by-the-book Scrum, a popular agile framework: Two-week iterations, story point estimates, and all the roles and ceremonies associated with it
The consulting company collects a fat check and leaves
Employees are unhappy with the overhead and heavy-handed processes, and efficiency does not, in fact, increase
The problem: Neither of these companies was a traditional software company. They were a research-first hardware company and a large "legacy" industrial company, respectively. Work there just does not fit neatly into two-week increments of small, estimable user stories. In the case of the former company, the fellow I talked to complained:
"Now I can't just go read a research paper. No, I have to write a user story first about what I'm researching. Then I have to give an estimate for how long it'll take me to read that paper, and every morning during standup, I have to say that I'm still working my way through the paper."
Doesn't that just sound like the opposite of agility?
In the case of the industrial company, the lament can be summarized as, "Everything we do is on a large scale with complex interlocking processes; nothing there can get done in two-week increments."
Now, with AI, many companies are in danger of repeating the mistake of using the wrong methodology to explore it, by going too wide too soon, and adopting a top-down mandate driven directly from the C-suite, supported by a one-size-fits-all playbook courtesy of the Big Expensive Consulting Co.
Companies would do well to remember Gall's Law, which states that anything complex that works must have gradually evolved from something simple that worked. This goes for adopting agile methodologies as much as it goes for integrating AI into the company. Small pilot, learn what's required for your company specifically to make it work, and don't expect much value from an off-the-shelf, by-the-book transformation, whether it's agile or AI.
