Lessons Learned From Basecamp Debacle

For a long time I’ve felt that we need a better way to do capitalism, or more generally a market based economic model. My first foray into actually studying that was running across John Abrams’s The Company We Keep back in 2007. I was suffering extreme burnout from starting my own company with the traditional workaholic model and thought there had to be a better way. In my own company at the time I started advocating for and pushing some of these sorts of things within the bounds that we could. If I ever started another company, I said, I wanted to do a radically different model of ownership, governance, and culture to really test out these newer ideas. In recent years I ran across the writings of the founders of Basecamp, David Heinemeir Hansson (DHH) and Jason Fried. They have a whole podcast called Rework which covers these topics as well. It is very much about rethinking a lot of the conventional wisdoms in the corporate world using techniques they’ve used at their own company for 20 years or so. I thought they were a good model to look at when trying to formulate what my own new corporate culture and structure should be. That fact is what made the events of the last two weeks even more distressing to me. I decided to re-read their two books It Doesn’t Have to Be Crazy at Work and Rework as well as looking at how these events unfolded to see how if I was confronted with the same scenario I would handle it differently. I’m not doing this as a beat up on Jason and DHH exercise but instead to try to process this to learn from this whole thing.

Problem Appearance at the Surface

Publicly the entire problem exploded into public view on April 26th with Jason’s blog post that they would no longer allow political conversations at work, would drop “paternalistic benefits”, drop peer annual reviews, and a few other things. The first one is what really set off the firestorm. It looked like the company was diving into the contrived “anti-woke” conservative messaging just like others have done. But this was supposed to be a company that has some sort of a moral compass, as much as any company could, and saw the folly of saying “work is just about work”. Then as the story continued to unfold people from the inside, including DHH himself, started chiming in with more details. For too many years Basecamp employees had maintained a list of “funny customer names” that they’d make fun of (see The Verge story on this). While many of the names being made fun of were of minorities not all of them were. A discussion forum on the internal site exploded into controversy on the topic especially once someone posted the Anti-Defamation League Pyramid of Hate. That caused the conversation to start spiraling out of control. In an attempt to de-escalate DHH posted his own commentary on the pyramid. That commentary proved to be an accelerant. That followed with an additional exchange where he allegedly dug up some past communications from another employee to use against them. That caused an HR incident against DHH which was dismissed. Then shortly after all that settled these new policies were suddenly announced in a blog post, being the first time anyone in the company even heard of much less saw that information.

Probable Actual Problem

In “It Doesn’t Have to Be…” they lay out the concept of something called a “trust battery” (pg 81) which is a concept they got from Tobias Lütke of Shopify. As laid out by Lütke:

Another concept we talk a lot about is something called a ‘trust battery’. It’s charged at 50 percent when people are first hired. And then every time you work with someone at the company, the trust battery between the two of you is either charged or discharged, based on things like whether you deliver what you promise

When a trust battery is low that’s when minor things like being a few seconds late for a meeting or a simple seeming question in an email can turn into full scale wars. When you have a combination of that low battery level with volatile events the situation becomes more dire. It seems that a big part of that which Jason and DHH were trying to address is changing the behaviors that were draining the trust battery rather than filling it. Their method of that though actually exacerbated the problem rather than quelling it. Why?

We need to look no further than the end of their “Trust Battery” chapter in “It Doesn’t Have to Be…” where they wrote:

Having a good relationship at work takes, err, work. The kind that can only begin once you’re honest about where you’re starting from. The worst thing you can do is pretend that inter-personal feelings don’t matter. That work should “just be about work.” That’s just ignorant. Humans are humans whether they are at home or work.

Their policy took this very common sense notion, which is sadly missing from many company cultures, and threw it out the window. They replaced it with, “You are just there to be another automaton in our personal software factory.” That was at least the message in the original post. They added more nuance later, while still not changing the policy though. More nuance in what they meant or not the damage had been done. The trust battery had gone from being drained to being split in two on the floor. Their clarification while better still left a bit to be desired as well. Why? There is some notion that needing to self censor is a new thing in the modern era. That before now everyone just sucked it up and dealt with it. It is the new era of the “liberal snowflake” or the “woke era” as labeled by conservatives and that media heads go along with that has led to these dire times. The truth though is that self censorship has happened forever it’s just that the ones self-censoring were everyone but straight white men. In this century the notion that they may have to self-censor to some extent too is what’s new. These “no politics” discussions are really about returning to the “before time” when they had the monopoly on no blow back for their words/actions and were the ones with almost exclusive seats in the positions of power to enforce that. You can’t put the toothpaste back in the tube though and to be frank that situation persisting as long as it did was ridiculous anyway. So these moves disproportionately affected the team members that were minorities which further broke the trust battery. However it gets worse than that.

One of the things highlighted in their books is the notion that “the owner’s word weigh a ton” (“It Doesn’t Have to Be…” pg 88). It’s literally a whole chapter of that book. When an owner makes a “suggestion” it is often interpreted as a command/directive. It is inevitable. As a former owner I had it happen all the time by accident. “Why did you all change that? I thought you wanted to go with the other design.” “Well Terry said you said it was a bad idea.” In actuality it was a passing pithy comment in a casual conversation by the coffee machines. This was in a corporate culture where when I put on a “team member” hat to help on a project rather than “Owner or CTO” hat I was explicit that the person running the project had final say and that my contributions should be treated equally with all others. It’s why one has to be judicious about how and when to weigh in on a topic. As another example even though I was horrified and upset about Trump winning the election in 2016 I was equally upset when I saw CEOs of some companies were sending out company wide emails saying anyone who voted for Trump is evil/not welcome etc. While I heard more than a few heated exchanges back and forth on the topic I felt me interjecting myself so directly and in that setting could be seen as creating a hostile work environment and I just sat that one out.

When DHH chimed into that very heated conversation about the ADL pyramid he should have known that was an escalation not deescalation, whatever his intentions were. I won’t get into the fallacies presented in the exchange snippet as published because it’s irrelevant. A better deescalation step would of been how he actually began: simply addressing the kernel no one seemed to disagree about and then putting a bow on that. “Making fun of our customers is wrong. It happening to racial minorities substantially is worse. Us letting it go on for decades is even worse still. We can and will do better and we need to make sure that some other festering wound like that isn’t still with us.” That’s it. Also maybe talking about wanting to take a further conversation about these things offline and to a media better suited for it like a telephone or video call maybe. Capping the discussion thread like that as a “we don’t need more discussion it was fucked up, we fucked up, we have to and will do better” would have been good deescalation. Throwing fuel on the fire, obscuring the original point, and then shutting it down after things accelerated further however just made things worse.

Lastly the formulation and roll out of their changes could have been done way better too. In Rework and “It Doesn’t Have to Be…” they had many sections that could have helped them here. “Don’t scar on the first cut” (Rework pg 260) warns against putting up bureaucratic responses that don’t need it like a dress code because a few people wore shorts work. “Why not nothing” (“It Doesn’t Have to Be…” pg 159) points out that, “It’s easier to fuck something up that is working well by changing it than it is to genuinely improve it.” A few chapters back on pg 139 “Don’t be a knee jerk” they point out how bad an idea it is to put something out there for everyone to have a knee jerk reaction to rather than a more detailed document meant to foster discussion. If there had been more deliberation in the policy change, along the lines of what they recommend in “Commitment, not Consensus” (“It Doesn’t Have to Be…” pg 152) a lot of the knee jerk reaction probably would have been abated. Supposedly the latter is their process but in this particular case at least it was the former. When it blew up in their faces they could have gone back to their advice in “How to say you’re sorry” (Rework pg 238) which was to not use non-apologies like “We are sorry if this upsets you” but instead true apologies. They aren’t talking about changing their policies though which makes me think they are thinking of their advice in “Take a deep breath” (Rework pg 244) which spends the first half talking about how change always pisses someone off and the pissed off ones will be most vocal so let time help settle things out. Again a lot of the churn here I think was from not following the other advice laid out but even so they should not forget the last part of that chapter which talks about engaging those upset to let them know you are genuinely listening and hearing their concerns but want to see how it plays out for a little while.

A Different Approach

Using all of this how would I have handled this differently? It’s easy to play Monday morning quarterback so I’m not trying to do that per se. I’m more thinking of this as a thought exercise. The root problem here seemed to be trust batteries and those wearing down over time. It is possible that constant bickering on message board systems are a big contributor to that. Was it always about political conversations or did this just highlight a larger problem in a very poignant example? If it is “always” around political conversations then figuring out a better method with the team should have been explored. The team are the ones who ultimately have to live this policy not Jason and DHH. If it is part of a larger miscommunication and trust battery issue that would be a different avenue of exploration. I’ve seen many times where people with adversarial relationships, i.e. low trust batteries, get into ever escalating email/IM wars when a simple phone call or walking over to the other’s desk for a face to face conversation would have instantly solved that. I’d probably want to explore that sort of thing being more explicitly stated.

It was stated by DHH that this was the second time this sort of over escalation of a political topic happened. Over what time period was that: a week, a year? An organization can’t have perpetual persistent dysfunctional communications but over correcting can be an even worse problem. Let’s go to a less charged hypothetical scenario where someone was found to be streaming porn on a company computer and network so people were reminded not to stream that sort of content or pirated music torrents etc. as is company policy. Let’s say it happens again with another person six months later who was streaming Wandavision episodes off of a torrent site. Obviously something needs to change since the message didn’t penetrate. However if I, the owner, single handedly decided and without notice that all network traffic besides to specific sanctioned sites would now be blocked, including all social media, YouTube, etc. you can be sure I just created a bigger problem than the original one. The same is true in this scenario. How do you create space for people to have real world discussions without making those that don’t want to be compelled to? How do you make sure that dysfunctional communication patterns that this is a bigger sign of are better addressed and how do you work to build back up depleted trust batteries? That would be the conversation with the team I’d want to have and move forward with. Ironically the suggestion of having it in alternative channels but at work may have been a great solution for the team but that wasn’t the original one and the whole messaging and lack of collaboration with the team block off receptivity to it.

Along those lines, knowing that a change on such a sensitive topic was going to be contentious even with all of the above techniques I wouldn’t introduce it with a whole host of other somewhat but not really related changes and then bundle the whole thing up with politically charged language that I should have realized would trigger even more adverse reactions to the already energized and upset team members. Imagine a scenario where the team is having lots of animated discussions and expressing being very upset about the company forcing way too many overtime hours and death marches for weeks and months on end. Then on a Friday the company puts out a “policy” change saying, “We won’t have overtime or death marches anymore we will have ‘added team contributions’ and ‘victory marches’, oh and we have a no crybaby policy since too many lazy people keep bitching.” It’s a wrong policy but also a wrong message. Imagine compounding that by adding on, “We’ve thought for a long time that our benefits policies were incorrect so now no benefits either.” Even if all that were true the timing and framing is just horrible.

Conclusion

It sounds trite but so often these sorts of conflicts fall down to failures of communication. We have that epically here in the lead up to the events, through the events, the formulation of how they wanted to deal with the events, and lastly their roll out. While most of their book of personal wisdoms on how they run a company don’t directly deal with this sort of subject they very much integratively do. Applying their advice to this problem I think there was a path forward that didn’t have to accept a false choice between a status quo where “for a second time political conversations got too heated” and radically changing things which would disproportionately silence their minority team members. It’s not just the initial policy but the process they arrived at it that could have been done better following the advice in their books. What is more I think their advice on underlying problems of “trust batteries” and dealing with that rather than the supposedly immediate problem could be a much more productive and important thing to address as well.



Picture of Me (Hank)

Categories

Updates (129)
Software Engineering (124)
Journal (119)
Daily Updates (84)
Commentary (73)
Methodology (59)

Archive

2021
2020
2019
2018
2017
2016
2015
2014
2013