We have very few rules at Atomic Object. Our employee handbook is intentionally labeled “Atomic Guidelines”, and is full of examples, expectations, and explanations but very few concrete rules. The rules we do have concern time tracking, start of the workday, and contributing to the blog. We favor guidelines over rules because it forces each one of us to engage with the spirit and intent of an issue, and leaves the responsibility for deciding on a course of action with the individual. (I am quite familiar with the folly of trying to create a perfect, complete specification, i.e. rules, for anything sufficiently complex, whether it’s a software system or a company.)

One bad apple is one bad apple

I’ll admit it, it’s hard not to make rules. When I’m frustrated with employees that consistently don’t meet our relatively few basic expectations, and I hear lame excuses and feigned confusion, I’m tempted to throw out our guidelines and make concrete, black-and-white, unambiguous rules with clear consequences for ignoring them. I step back from this cliff by reminding myself that we employ 47 people and only a very, very few push me in this direction. That’s the first lesson I want to share:

Don’t take general action to address problems with a few specific people.

I learned this one the hard way, unfortunately. I’ve discovered that a few problematic people can unreasonably poison my attitude towards a lot of people if I’m not careful. I try to remind myself, when I’m feeling frustrated, that for every problem I’m dealing with, I can find 10, 20, 30, even 40 people who are positive counter-examples.

Strongly resist generalizing from one to many when it comes to bad behavior.

Rules and games

Should you succumb to your frustration and create a system of rules, you will have created a game. This guarantees two things: 1) people will play the game, and 2) you will lose. I suspect that something happens when you create a lot of rules that distracts people and veils the original issue behind the rules. By shifting people’s focus to the rules, they lose focus on the original intent. Their creativity, time, and energy goes into playing the game, not achieving what the rules were intended to enforce. In this state, they probably start to feel more morally obligated to playing the game fairly than working in the business. At least that’s how my inner pop-psychologist explains it. Whatever’s going on here, the lesson I’ve learned is:

When you make rules you create a game, and people will play it.

The odds of an employer winning this game are small. To create rules, you’re forced to abstract a lot of the messy world into a simple model devoid of detail. That model will be found wanting when it meets the real world. There will always be cases you didn’t consider, situations you didn’t anticipate, subtleties and nuances you neglected. You will find your rules wanting, and someone playing the game will find plenty of ways around them. If you react by adding more rules, you’re entering into a cycle from which you can’t escape. Your hole is getting deeper, and you’re still digging. Even if you persevere to an end-state where you achieve perfect compliance with the rules, you lose. You’ve wasted a lot of time and energy in playing the game. You’ve probably pushed employees into finding other ways of “beating” you. My conclusion is that:

You can only lose such games.

Consequences

If you acknowledge the folly of rules, what happens when you have guidelines that are clearly communicated and well-understood, and one or more people choose to consistently ignore them or not act in good faith with them? Avoiding rules doesn’t mean not having consequences. Unfortunately, it’s been hard for me to find ways of defining consequences that aren’t full of rules. I think this comes from wanting to be fair.

If the consequences have teeth, it’s only fair that people know what they are, and when they’ll be triggered. That reasonable statement leads into the jungle of rules, unfortunately, but having no consequences for ignoring important guidelines is undisciplined, and unfair to the people who share the pain, pull their weight, and own their responsibilities.

I’ve realized, thankfully not from experience, that the “nuclear” option is useless in most cases. If you’re frustrated and angry, it’s tempting to think, “Screw this; I’m sick of putting up with this. If you don’t comply, you’ll be fired.” That’s a nearly useless consequence, as you’re very unlikely to ever pull the trigger on it. It’s an idle threat for anything other than the most egregious and serious failures of responsibility.

The nuclear option (firing) is useless.

Effective consequences, and their enforcement, have these common characteristics:

• Are aligned with the negative effect of having ignored the guideline or expectation.
• Provide a path back to compliance and re-alignment with the company.
• Have an element of mystery or uncertainty to them, without being arbitrary or capricious.
• Are judged by a human, not an algorithm.
• Are open and have a component of social judgment.

An example: blogging

Atomic’s blog, Spin, is a shared company blog that everyone at Atomic contributes to. Our approach makes for a rich, authentic, and diverse set of topics and posts, and is consistent with our value mantras, in particular, “share the pain”, “teach and learn”, and “own it”.

We came to this approach after earlier failures, plenty of debate, and education. No one who lived through that time period could fail to remember or appreciate the arguments or understand the expectation for blogging. Those expectations, and distillation of the rationale for our blog, are thoroughly documented in our guidelines wiki. The vast majority of the company writes excellent posts and meets the basic expectation that everyone will contribute a post at least once every 40 days (we’ve adjusted that number up as the company has grown– from 30 days, two years ago, to roughly match the supply of posts to a once-per-day publishing schedule).

A few people, unfortunately, are prone to miss their deadlines, offer unacceptably short, trivial posts or enter sloppy, rambling, error-filled drafts to technically comply with their every-40-days deadline. (Thankfully, these don’t make it to Spin, as each post goes through our expert and patient editor, Lisa Tjapkes.)

The consequences for “going red” (we dedicate a portion of our central office information radiator to showing everyone’s blog status) have been nothing more than a guilt trip: explain to our Board through reply to an automatically generated email why you failed on your commitment. Responses to this consequence have ranged from contrition and regret to completely ignoring the email to the functional equivalent of “I didn’t feel like it.”

Clearly, at least for some people, there are no teeth in this bite. When letting this situation ride was no longer acceptable to me, or fair to the vast majority who make the effort to blog, even when it’s not fun or perfectly timed, I decided new consequences with more teeth were necessary. I tried to follow the guidelines above when I devised them.

New consequences

I made no change to the guidelines: blog about what you like, do so at least every 40 days. You’ll still get automated emails when you’re 10 days from due, and when you’ve gone past due (turned red). You’ll still be expected to explain yourself to the Board. But you may also lose 50% of your cash quarterly bonus for the next quarter and be suspended from participating in Spin. In addition, if you are suspended from Spin, the days between posts for everyone else will be decremented to ensure we have enough content.

There are no hard-and-fast rules about when the suspension will be triggered (that would create a game). There are no documented edicts like “one goof per person per year” (more games). A human will judge each situation (they are rare enough to do that, and hopefully will become even more rare) so that legitimate excuses like family emergencies or sickness don’t trigger suspension. After a quarter of suspension, you may re-apply to participate in Spin and regain your full quarterly bonus (there’s a path back to a place of integrity). The forfeited bonus will go towards our other marketing efforts (the punishment is aligned with the negative consequences to the company). Your individual line on the information radiator will turn black to mark your suspended status (social consequences).

If I were to ignore what I’ve learned about these situations, the new consequences would be much different: You get one mulligan per year. When you go red, the trigger is pulled. The punishment is being fired. No one else is aware of your fail, or feels any pain. It’s a design that violates every lesson learned, and yet, wouldn’t be that difficult to talk yourself into when feeling well and truly frustrated.

I hope our new design reduces or even eliminates the failures of those who don’t meet our expectations. It will also be interesting to see if this new approach has any unintended consequences. I’ll write again when I have some data to share.