4 ms·
A real life example in closed source development in a tech company that I experienced directly... A team organically chose a cute mascot animal and started doi
by rented_mule 2y ago
A real life example in closed source development in a tech company that I experienced directly...
A team organically chose a cute mascot animal and started doing things like bringing in stuffed animals of that animal and including visual puns involving that animal in presentations. The manager leaned into it, which caused people to find even more ways to have fun with it. As the team grew and split, it became the mascot of the group of teams. Some engineers got together and wrote a short graphic novel explaining the emergence of the mascot in the group. Eventually one of the engineers designed a logo for the group that was a cartoon version of the animal and subtly incorporated the group name into it (the name was almost an Easter egg that had to be discovered, it was very clever). Stickers of the logo were made and given to new engineers when they joined the group. People would use "where did you put your stickers?" as an ice-breaker to have a first conversation with new people. Reserved people who were hesitant about giving presentations were encouraged by getting laughs from their visual puns. It was all very silly, but it made a lot of people feel included in the group soon after joining, opening connections and fostering better communication / collaboration. Highly tangible benefits were easy to spot. This lasted for the better part of a decade.
When the leader of the group (the original manager that leaned into it) left that role, the new leader viewed all the silliness as a distraction. He didn't explicitly kill it, but he did lean away from it. He never included any references to it in his presentations. When asked to reimburse a batch of stickers for new hires, he said the company can't waste money on things like that (less than a dollar per new hire). He spent hours removing all references, usually in the form of textual puns, from documentation. In a year or so, the mascot was relegated to nostalgia. New people were no longer meeting as many people in their first week and cohesion suffered. There was less fun in general. It was a big part of the group's culture becoming much more sterile and detached. When asked about it, the new leader proudly called that sterility "professionalism". A lot of people that had been in the group for many years called it "one less reason to stay".
Years after the mascot faded away, there are still group dinners of ten or more former employees of that group, that happen multiple times a year. A recurring topic of discussion at those dinners is how to find a place to work that is so motivating and has such great teamwork. The silliness and simple jokes around that mascot were a big part of what made that group so cohesive. In fact, they were both a symptom and a cause, creating a positive feedback loop that helped build and maintain the group's culture.
- vintagedave 2y ago> called that sterility "professionalism" Bingo. Off topic, but this is what bothers me with current macOS or Windows UI design: it is sterile. It's not nearly as fun as, say, Windows 7 (or XP!) or pre-flat macOS. It's very easy to move something from fun, vibrant, exciting to flat, boring, and sterile. It is _very hard_ to move it the other way, because of concerns about professionalism, where professionalism is seen as sterility and the safety that's in that sterility. I think UX needs _joy_... would you agree, are you and I and the parent poster are saying something similar about different aspects of computing / software? I wrote an article on it here: https://daveon.design/creating-joy-in-the-user-experience.html https://daveon.design/creating-joy-in-the-user-experience.ht...
- rented_mule 2y agoYeah, I agree that this is all related. Especially the safety part. All of the overlapping domains we're talking about here are at their best when creativity is unleashed. Sometimes that means failed experiments (uh oh, danger!). But it's also where we get big leaps forward. Managed properly (e.g., doing the hard work of learning from and addressing failures before they do too much damage) can lead to a much better average over time. I think this is all analogous to the conditions that give us bland TV commercial jingles vs. the conditions that give us great music.