The engineering mindset | Will Larson (Carta, Stripe, Uber, Calm, Digg)

Jan 7, 2024 Episode Page ↗
Overview

Will Larson, CTO at Carta and author, discusses evolving engineering leadership, strategy development, improving EM-PM relationships, and effective ways to measure engineering productivity. He also shares insights on finding time to write and defining company values.

At a Glance
15 Insights
1h 16m Duration
14 Topics
3 Concepts

Deep Dive Analysis

Changes in the Engineering Landscape and Market

Treating Engineers as Adults and Leaders

Understanding and Applying Systems Thinking

Developing Effective Engineering Strategy

Learning to Get Good at Strategy

The Importance of Writing for Career Growth

Tactics for Making Time to Write

Tips for Aspiring Writers and Content Creators

Fostering Productive Relationships Between Product and Engineering Managers

Aligning EM and PM Performance Incentives

Measuring and Improving Engineering Productivity

Defining and Applying Company Values Effectively

The Digg v4 Rewrite Failure Story and Lessons Learned

Will Larson's Upcoming Book and Key Takeaways

Systems Thinking

A method of analysis that involves thinking about 'stocks' (things that accumulate, like the number of fish in a lake) and 'flows' (the movement between stocks, such as fish being caught or reproducing). It helps model reality to identify conflicts between a mental model and actual outcomes, providing opportunities to learn and improve the model and subsequent actions.

Good Strategy (Richard Rumelt)

A strategy defined by three core components: a clear diagnosis of the current status quo, guiding policies derived from that diagnosis to address the situation, and specific actions to implement those policies. Good strategies often involve setting constraints to focus limited resources on the most critical problems, rather than trying to please everyone.

Identity Values

Company values that are universally positive and non-reversible, meaning it's hard to imagine a company operating with the opposite value (e.g., 'we build good software' or 'we care about our customers'). While not inherently wrong, these values are often not useful for guiding specific decisions or serving as a hiring filter because they lack practical differentiation or actionable guidance.

?
How has the engineering market changed recently?

The market has shifted from a 'zero interest rate era' of rapid hiring and growth to a period of slower hiring, team consolidation, and increased focus on efficiency and accountability for engineering leaders, requiring them to lead teams and manage resource allocation more directly.

?
What is systems thinking and how can it be applied?

Systems thinking involves modeling problems using 'stocks' (accumulations) and 'flows' (movements) to understand how different elements interact over time. It can be applied to areas like hiring pipelines to identify bottlenecks and prioritize improvements by comparing the model to real-world data.

?
What makes a good engineering strategy?

A good engineering strategy, as defined by Richard Rumelt, includes a clear diagnosis of the current situation, guiding policies to address it, and concrete actions. It often involves setting constraints (e.g., using only existing tools or a specific tech stack) to focus energy on core business problems rather than appeasing everyone.

?
How can product managers build more productive relationships with engineering managers?

Product managers should focus on deeply understanding their engineering manager's true needs and incentives, rather than assuming negative intent. Aligning performance incentives, such as giving EM-PM pairs the same performance rating, can also foster shared accountability and better collaboration.

?
How should companies measure engineering productivity?

Instead of relying solely on mechanical benchmarks or easily manipulated metrics, leaders should talk to engineers directly to understand effectiveness. They should also align engineering evaluation with business and product goals, and transparently show the roadmap of valuable, impactful work. Imperfect metrics like DORA can be a starting point for diagnosis and educating stakeholders.

?
What are the key characteristics of effective company values?

Effective company values should be honest (reflecting actual practices), applicable (guiding real-world decisions), and reversible (meaning you can imagine a company operating with the opposite value). They should avoid being merely 'identity values' that are universally positive but offer no practical guidance or differentiation.

?
What was the 'Digg v4 rewrite' failure and what was learned from it?

The Digg v4 rewrite was a complete overhaul of the platform aimed at adding social functionality, which resulted in a month-long period of the site being largely non-functional due to a core bug. Despite the business ultimately failing, the experience provided invaluable learning opportunities for the engineers involved, shaping their careers by forcing them to solve critical problems under extreme pressure.

1. Treat Engineers as Peers

Treat engineers like peers and put them into senior leadership roles, holding them accountable, rather than coddling them, to foster their growth and allow them to tackle real problems.

2. Use Systems Thinking

Apply systems thinking by modeling problems with “stocks” (things that accumulate) and “flows” (movement between stocks) to understand reality, identify conflicts with your mental model, and pinpoint where to focus efforts for improvement.

3. Write Down Your Strategy

Document your strategy (for any function, including engineering) even if it’s imperfect, because a written strategy allows for debugging, improvement, and consistent application across the organization.

4. Embrace Strategic Constraints

Implement “boring” strategies that constrain options (e.g., using only existing tools, specific tech stacks, or data centers) to focus energy on company-valued problems and align teams, even if they are initially annoying to some engineers.

5. Align EM-PM Performance

To address misaligned incentives and foster better collaboration, consider giving Engineering Managers and Product Managers the same performance review rating, reinforcing that they are a “pair” with shared accountability for overall execution.

6. Understand Others’ Needs

Before trying to solve a conflict with an EM or PM, dig deeper to understand the other person’s true needs and what they genuinely care about, as often a compromise can be found that satisfies everyone without extra time.

7. Measure Engineering Productivity

Combine benchmarks and metrics like Dora (useful for diagnosis) with qualitative insights from engineers, and align engineering evaluation with business and product goals. Be comfortable measuring imperfect metrics and use them to educate stakeholders on nuances.

8. Write What Energizes You

To sustain a long-term writing habit, prioritize writing about topics that genuinely energize you and that you are curious about, rather than focusing on what you think others want or what’s popular, to avoid burnout.

9. Align Writing with Work

Find ways to write about topics directly related to your work, as this allows you to refine your thinking, improve job performance, and integrate writing into your demanding schedule rather than treating it as a separate, distracting activity.

10. Publish Consistently

If your goal is consistent long-term writing, prioritize publishing almost everything you write, rather than getting stuck on perfecting drafts. Accept that some content won’t be “perfect” and focus on sharing your evolving thoughts and experience.

11. Define Values Honestly

When defining company values, ensure they are honest (you actually do what you claim), applicable (you can use them to guide decisions), and reversible (the opposite could also be a valid choice for another company), to make them truly useful.

12. Focus on Quality Artifacts

If your primary goal for writing is career advancement, focus on producing two or three exceptionally good pieces, dedicating significant time to drafting, revising, and getting feedback, rather than trying to maintain a frequent publishing schedule.

13. Don’t Stress Minor Decisions

Before agonizing over a decision, ask yourself if anyone will remember or care about it in six months. If the answer is no, make a reasonable choice and move on to more important matters.

14. Ask Candidates About Choices

When interviewing, ask strong candidates how they plan to decide between multiple compelling offers to understand their priorities. Then, tailor your pitch to highlight how your company best meets those specific needs.

15. No Way Around, Just Through

In challenging situations, adopt the mindset that there is “no way around, just through,” meaning you must confront and work through difficulties directly to reach the other side, rather than trying to avoid or dodge them.

I think that we often treat engineers a little bit like children instead of giving them like the responsibilities and ability to actually thrive as adults.

Will Larson

Reality is always right. Your model is always wrong if it's in conflict with reality.

Will Larson

The biggest risk to content creation of any resort is quitting soon because you get burned out.

Will Larson

There are no villains in the workplace. They're just people with like complex incentives that are doing complex things.

Will Larson

Metrics are about educating the people consuming the metrics about the reality of the rich data underneath.

Will Larson

Will anyone remember what we decided in six months? So I think people stress out about a lot of decisions. But I increasingly believe like most decisions people stress out about just like aren't that important.

Will Larson

Strategy Definition (Richard Rumelt's Components)

Will Larson (attributing Richard Rumelt)
  1. Diagnosis: Understand the current status quo and real-world constraints.
  2. Guiding Policies: Based on the diagnosis, define how to address the situation.
  3. Actions: Implement specific steps to execute the guiding policies.

Will Larson's Writing Process for Books

Will Larson
  1. Write the entire book before engaging with a publisher.
  2. Diligently anticipate publisher concerns to ensure content reusability.

Evaluating Company Values (Will Larson's Criteria)

Will Larson
  1. Honesty: Do you actually do what the value claims?
  2. Applicability: Can you apply this value to make real-world decisions?
  3. Reversibility: Can you imagine a company operating with the opposite value?
roughly every four months
Layoff frequency at Yahoo (early career) During Will Larson's early career at Yahoo.
up to six
Interviews per day at Uber (peak hiring) Back-to-back interviews for engineering managers.
two or less, possibly zero
Current interviews per month for engineering managers Reflects the shift in the engineering market.
three months
Time to spin up Uber China data centers Achieved due to Uber's 'no cloud' strategy.
approximately 1,000
Number of Will Larson's blog posts Over 16 years of writing.
approximately one month
Digg v4 site functionality restoration time Period during which the site was not working properly after the rewrite launch.
every 12 hours
Digg v4 server restart frequency (initial fix) Required for every server even with a caching mechanism, before the core bug was found.
from ~100 to ~30 people
Digg team size reduction after layoffs Occurred within nine months of Will Larson joining Digg.