Week 45, 2024 - Launch Experiences

Important week in the US, but I don’t think I want to bring politics into this newsletter, regardless of how the events could shape our personal and professional lives.

Side note: I disagree with the movement of ignoring news – but I strive to select when, how much, and what I’m reading. Rolf Dobelli has many good points, especially if you apply them to a specific kind of news and the doom-scrolling their consumption can induce. I try to search for content that’s rooted in my beliefs, but still challenges them somewhat. This careful extension of the comfort zone is where learning can happen — and I find that this approach works for news consumption too.

So, without any deeper political context, here’s a site that I, coming from a country with GDPR's fair level of privacy regulations, can’t believe legally exists: VoteRef.com. If you live in the US (or use a VPN), you can search for individuals who voted in any of the recent elections by name or address, and list a lot of their personal data, including age, people they share a household with, when and how they voted*, etc. I found a few of my friends and felt super uncomfortable seeing data I believe should be private. I am a big fan of public data, but it feels to me that this is going a bit too far. (Paywalled source article here.)

*"how they voted" refers to mail, in-person, etc., not who they voted for, though party affiliation, meaning registration to be able to vote at the primaries is shown. Thanks for the clarification, Mark!

📑 What I learned this week

Jeremy and I launched the new Ghost-based site of our podcast, The Retrospective. There is still a lot we want to improve, but at this point, we have reached a “good enough” point and would rather do the remaining changes live. I’m super excited about the direction our show is taking by being better focused and organized.

We recorded the 4th episode of Season 2 (to be published next Tuesday), and used a different approach in preparation: instead of just listing high-level bullet points, Jeremy worked on a script to use as guidance, and we spent some time with it before recording to finalize the structure. I was afraid this would remove some spontaneity, but in reality, it was a great help to ensure we’re aligned on what we want to say, how, and when. As a result, I think we created one of our most organized episodes yet. You be the judge of it - add your email here to get a notification when it’s live!

And a mini-learning: having a popular URL in your LinkedIn post decreases its visibility. I guess LinkedIn is trying to ensure novel content is being shown to people – regardless of personal commentary.

🤔 Articles that made me think

The 2024 DORA Report

I wanted to cover the annual report it this section, however, the content grew over what I believe is appropriate for these Friday updates, so I wrote a separate article listing what stood out for me. In short: mostly positive impact of AI and Platform Engineering, with some interesting drawbacks, and important findings about Developer Experience. Read my takes here.

Platform vs. DevEx teams: What’s the difference?

Focusing on just the difference by surveying various Platform and DevEx teams. Instead of trying to come up with a definitive differentiator, the post concentrates on the differences and similarities in various approaches of different-sized companies. If I had to simplify, it seems that DevEx teams’ focus is rather on the “Build” stage, and Platform teams have more remit in the “Operate” area, while both have “Enabling” close to their mission.

Fixing Employee Performance Management (Again?)

Great article on the subject with a good historical background, and a screenshot-filled analysis of the actual HR tool used at Revolut. Interesting takeaways for me:

• History is repeating itself: employee performance management is an ever-swinging pendulum between accountability- and development-focused approaches.
• At Revolut, the department responsible for performance management is separate from HR, and reports directly to the CEO. Building a high-performance organization is the CEO’s job.
• There’s no one golden performance management method, it hugely depends on the organization’s context. Optimizing more towards an accountability-based approach makes sense for a pre-IPO company that’s trimming down the remains of previous over-hiring. Then, focusing more on development could be a sensible approach for the same kind of company a few years later, struggling with employee morale and retention. Just like most things with leadership: it depends.

(In case you’ve missed it back then, I wrote about the topic with a rather development-focused approach in mind here: Engineering Managers' Guide to Effective Annual Feedback.)

Saving $10MM in 5 years by leaving the cloud

I’m sure you remember when DHH announced that the apps of 37signals (Basecamp, HEY, etc.) were planning to leave the cloud. (Even creating self-hostable, anti-SaaS alternatives under ONCE.) Going against the trends is always interesting, and it’s great when we get a follow-up. This time it seems that despite high initial capital expenditures, navigating around long-term contract commitments, and increased operational complexity, the savings can still be significant. Another lesson filed under “it depends”.

That’s it for today, get the bigger context this weekend,

Péter