Dents vs. Rats

At the Microsoft Build analyst day, Julia Liuson (VP of Corporate Tools) spoke about Microsoft’s journey to building more customer-centric products. A company’s history impacts their culture; the activity of delivering generational software updates in the 1990s remained evident in some of Microsoft’s processes and internal language decades later. Luison shared lessons the company made in adapting their culture to better align with current software delivery trends.

Part of this journey involved internal examination. If language affects actions and actions reflect culture, then it follows that how people talk about things can impact how teams operate. Did it make sense to continue using terminology like “recall class bugs” or “RTM” (release to manufacturing) in the internet age of software delivery, where bugs can be fixed with updates and physical manufacturing is no longer a gating part of the software delivery process? Probably not. What about “back of the box” features? Teams used to prize when the features they worked on were important enough to be listed on the back of the box, but perhaps that language was limiting. Instead of focusing on big features, Luison wanted teams that were willing to iterate and work on small, incremental changes.

My take on Julia Liuson's talk at #MSBuild #MSARC: Humility is key to customer centricity. Need to shift the culture from one of telling to one of listening. Or, in her words, knowing to learning. Great to hear how they did this @Microsoft Dev Tools division. pic.twitter.com/gZiWQv4CUZ

— Sam Charrington (@samcharrington) May 5, 2019

The journey also involved examining how they incorporated external feedback. My favorite part of this process was the importance of qualifying user feedback responses with the metaphor of “dents vs. rats.”

The metaphor began with a show-of-hands question polling the room about how many of us had dents in our car. A decent number of attendees raised their hands. (Shout out to myself that I was not one of them! woot!) Anyway, a fair number of people raised their hand, and I estimate that a majority of people kept their hands up when asked how many cars had been dented for longer than a year.

Luison hypothesized that if you polled the group about what annoyed everyone most about their vehicles, a large proportion of respondents would answer the dent in their cars was the highest priority issue. However, based on their behavior over the past year, this answer would be disproportionately representative of a problem that most users didn’t find actionable. Dents can be annoyances but are otherwise problems that don’t hinder activity or use.

On the other hand, Luison offered this hypothetical: “what would you do if you found a rat in your house?”

My answer would be to nuke the place (as those who listened to the Hark podcast episode where Steve asked me about my most feared animal may have guessed). This is maybe an extreme reaction, but even the more RATionally minded amongst us agreed it would be a situation that required immediate action. Some people might call an exterminator and some might handle it themselves, but few if any would do nothing.

This is the difference between dents and rats. Some problems are irritations and annoyances; it is better to avoid dents when you can, but also they are problems that many people can live with at least temporarily. Other problems, however, will immediately impact someone’s behavior until it’s resolved. Resolution to rat problems will involve consuming efforts to find a fix (either with in-house or third party solutions) or, for the rat paranoid amongst us, discontinued use of the product all together.

There are a few lessons Microsoft imparted with this metaphor:

1. When you ask for feedback try to take the time to understand the users’ associated behavior, not just their description of what they assume is the desired solution. The behavior attached to the problem is a key indicator of whether the problem is a dent or a rat. (Sometimes you’ll even discover the features they ask for won’t actually fix their problem.)
2. Prioritize fixing rats first. Dents may cause annoyance, but rats can be deal breakers.
3. Limit the distance between your product team and your customer to better understand user behaviors and how to prioritize your engineering efforts.

---

Disclaimer: Microsoft is a client and paid for my T&E for Build.