How do you spot patterns?

I enjoyed Melissa Fisher’s (@melissafisher) post on patterns.

She captured several handy ones:

  • Tight deadlines causing people to rush. Thus mistakes are made (we’re human, mistakes are part of the game). As a result problems are introduced into the solution.
  • A lack of communication and collaboration amongst disciplines. Therefore people being on the wrong page and building the wrong thing or taking extra time to build.
  • Ineffective meetings like sprint refinements. Requirements that are not clear, not testable and even gaps. Thus introducing problems.
  • A lack of understanding about what testing is about, which results in pushes to reduce the testing. Thus resulting in again problems in the solution.

I’m curious how as we progress in our careers we end up spotting patterns.

Is there a way to accelerate pattern spotting?

What techniques do you use to spot patterns?

How do you get to that stage where you can notice a pattern without leaning too much on a historical bias – and falling into the risk of making assumptions/clouding judgment?


Saw an interesting point about how leaning on our past experience can be a distraction.

  • We all gather data to tell us if a process change or product feature works. Many teams like to wait too long, to see how their work has impacted effectiveness long term, before changing something else in the system. Instead, agree on what a “win” looks like, and the moment you get the benefit of that feedback jump to the next optimisation.

I follow a data scientist who advises on clinical trials. You carefully calculate things like dosages and if the drug works you do the same to the next trialist until you have proof. But some people like to wait long after the patient is technically cured to find out if maintenance doses of a drug to use long term are correct. This even after doing all the modelling work to determine the optimum dosage very meticulously. Sometimes in corner cases they aren’t exact, statistically insignificant cases, which will be ignored anyway. But waiting to find out even if you know that you will ignore it if even 1% of the patients need a higher maintenance dose can cost money. We use past experience far too often to inform risk.

Another one is when managers seem to get stuck on delivering any kind of answer on a strategy topic. We have to assume that managers “do not know” and really come to accept that not knowing for example, what is in the roadmap is just how things are. And that we should all just agree that the direction is perhaps rubbish, and instead just get on with the job at hand. I see this continuously when companies present revenue growth, 90% of the time they don’t know with any certainty what drove any growth or shrinkage.

  • I often feel like I’m in a rowing team, and the CEO is sitting at the front of the boat and glancing ahead every so often over the shoulder to alter course. Of course the captain keeps shouting course corrections, it’s our job to make sure we execute those corrections as quickly as possible. It’s also our job to make sure that we row as efficiently as possible so we have good speed, that’s not the captains job at all. It’s not our job to question the zig zagging either, and neither to be dependant only on the captain for all course corrections too.

Many teams never build that communication and trust in their team as a whole. Management guru Peter Druker: “Only three things happen naturally in organisations: friction, confusion, and underperformance. Everything else requires leadership.”

Tight deadlines or “iron triangle” thinking are in my mind the boss trying to remind us that every project is a time-bound experiment, designed to earn him/her more money, nothing less, nothing more. So finishing early is of the essence, they do not have a long game for this play. Once you have some ability to step back a lot of patterns emerge, but observing from far, takes a lot longer.

1 Like

For spotting patterns I use a simple heuristic:

Once is happenstance, twice is coincidence, three times is a pattern.

When I suspect a pattern, I look at what happens before and after the suspected pattern, to see if the data supports my hypothesis. Then, if I think I am seeing a pattern, I look for potential causes and effects. Once I’m reasonably sure of a pattern be it good or bad, I choose an appropriate action.

Currently, the pattern I’m seeing is that I’m going to be without employment once the conversion project with my current job is over. There are no open positions and I see no evidence that there is any intention of moving me into any of the company’s existing departments. My action is to step up my job searching activities and try to update & improve my resume. I’m contracted until the end of March, but who knows what happens beyond that?

Assuming nothing changes, I’ll be sending out applications by mid-February and hopefully will have something secured when the time runs out.

Maybe it’s just paranoia, but I’ve found when similar circumstances happen three times, it’s a pattern and needs to be handled in some way.