Roles
In order to encourage learning and iteration, data science roles need to be made more general, with broad responsibilities agnostic to technical function.
Unpopular Opinion - Data Scientists Should be More End-to-End
You may not be end-to-end now. That’s okay—few people are. Nonetheless, consider its benefits and stretching closer towards it.
What’s in a name?: The semantics of Science at Lyft
- Lyft outlines the problems in data title ambiguity and proliferation.
The Analytics Engineering Guide
One Data Science Job Doesn’t Fit All
Lessons from Keith Rabois Essay 3: How to be an Effective Executive
Teams
Culture
It’s not perfect. We ship too many features, many half-baked. The product is complex, with many blind alleys. It’s hard to integrate non-engineers – they aren’t valued. But, we ship.
Engineers Shouldn’t Write ETL: A Guide to Building a High Functioning Data Science Department
Everybody Wants to be the “Thinker”.
Building a data team at a mid-stage startup: a short story
Wow! This is depressing! Let’s talk about what you can actually do to break out of this.
Data as a Product vs. Data as a Service
The difference between providing “data” and providing “insights” (actually though).
Run Your Data Team Like A Product Team
Service-oriented data teams aren’t effective.
If a human operator needs to touch your system during normal operations, you have a bug.
DataOps Principles: How Startups Do Data The Right Way
If it’s someone’s job to write a new SQL query or download data from external systems to handle all data requests, your team is headed in the wrong direction.
Building The Analytics Team At Wish
In this post, I’m going to share the lessons we learned, and offer a roadmap for other companies looking to scale their analytics function.
The Problem With Hands-Off Analytics
To put it another way, the only things self-serve helps scale is SQL…
Analytics is a mess: You can’t stop it, and you shouldn’t try to contain it.
When approaching any analytical problem—from something small like answering a single question to introducing a data practice at a company that’s never had one—we should expect the first steps to be uneven and uncomfortable.
Every senior person in an organisation should be aware of the less glamorous - and often less-promotable - work that needs to happen to make a team successful.
Communication
Super Specific Feedback: How to give actionable feedback on work output
By giving feedback on the work product, you’re giving feedback on the strategic thinking that went into that asset.
If you work on anything worthwhile, sooner or later people will care about it and will want you to send progress updates.
- You must communicate your work well. Start by sending good progress updates.
Please don’t say just hello in chat
Imagine calling someone on the phone, going hello! then putting them on hold…
The XY problem is asking about your attempted solution rather than your actual problem. This leads to enormous amounts of wasted time and energy, both on the part of people asking for help, and on the part of those providing help.
Responses to Negative Data: Four Senior Leadership Archetypes
If you are in the data business – my bread, butter and tofu – you often carry the burden of being the bearer of bad news.
Peer Review
Strengthening Products and Teams with Technical Design Reviews
How to review an analytics pull request
The Art of Giving and Receiving Code Reviews (Gracefully)
How to Do Code Reviews Like a Human
Unlearning toxic behaviors in a code review culture
Google’s Code Review Guidelines
Prioritization
One thing that always bugs me about (many) prioritization conversations is that teams often leave out the expected value curve of the work.
Shape Up: Stop Running in Circles and Ship Work that Matters
- A nice guide for breaking away from traditional, broken planning cycles.
Prioritizing Data Science Work
Answering one question with data often leads to new questions, so fulfilling requests often creates additional work rather than lowering the amount of work left to do.
The Tool that Will Help You Choose Better Product Ideas
- I’ve found that small prioritization frameworks like this greatly improve the focus of teams.
Most Data Work Seems Fundamentally Worthless
There is a flavor of despair I’ve become accustomed to, so deeply ingrained in the hearts of myself and my colleagues that it has settled into a hopeless passivity. It’s the despair that comes from knowing that we spend most of our time producing nothing of value.
Organizations
Culture
Let’s say you were employed as a CTO behind the front lines and you wanted to destroy productivity for as long as you can without getting caught. You can of course make a series of obviously bad decisions, but you’d get fired quickly. The real goal here is to sap the company of its productivity slowly, while maintaining a façade of plausibility and normalcy. What are some things you can do?
Why is culture so important to a business? Here is a simple way to frame it. The stronger the culture, the less corporate process a company needs. When the culture is strong, you can trust everyone to do the right thing.
There is room up in organizations to boost performance by amping up the pace and intensity.
- The history of Netflix’s culture is a great case study – they did the opposite of everyone else through very late stages of growth.
Coordination Headwind: How Organizations Are Like Slime Molds
An approachable presentation, done in an emoji style, showing why even when everyone is competent and collaborative, you can still get hurricane-force headwinds.
Structure
Mo’ Titles, Mo’ Problems
Models for integrating data science teams within organizations
- A comparison of the popular models of integrating data science teams within companies.
How to structure your data team
We figured it would be most useful to collect a bunch of “reference architectures” from amazing companies.
In short, the output of your organization is dependent on the number of people that can own projects and see them through to the end.