SWLW #507: Destructive metrics, The Truth Curve and the Build Curve, and more.

Heya,

As always, below you can read my best findings for the week. If you want to support my work or join our SWLW community in Slack - scroll down to the bottom of the email to see how.
 

This Week's Favorite

​​Use One Big Server
8 minutes read.

"Tall is Better than Wide" and "Use the Cloud, but don’t be too Cloudy" -- Beautiful framing to consider. I love this post because it makes you look around and see where else we got confused about complexity. One of the questions I love asking in interviews is "what's getting harder when moving to microservices" to see if people have explored the downsides enough. It's always a tradeoff, but we need to be balanced in our perception. Where else might it imply? Smaller teams (3-4) or bigger teams (6-9) can be one example. Look around you. Did you move the complexity one level (or more) up, thinking it would make things simpler? It's always good to review our decisions and see if we now understand the context in which they fit best.

Read it later via Pocket or Instapaper.
Share it via Twitter or email.

Product [sponsored]

 Promote your product on SWLW and reach over 30,000 leaders 

Culture

When You Have to Go Down Just One More Layer to Debug
1 minute read.

My humble effort to help you start the weekend with a smile on your face.

Read it later via Pocket or Instapaper.
Share it via Twitter or email.


KPIs, Velocity, and Other Destructive Metrics
7 minutes read.

This is one of the most important posts for people who run organizations as managers. You have to read it carefully without fighting the urge to reply, "well, we have to plan somehow," as you'll miss the point. I think that Allen Holub's examples for better KPIs also fall short in other aspects (not every outcome is interesting to measure, e.g. TDD for seed-stage company), but that's precisely the point. Practicing choosing the right KPIs for your organization will help you make sure they track outcomes you care about and care to optimize. Before you write your KPIs, write a few words about the outcomes you care about in the next 12-18 months and explain the reasons. Why is it important? Why now?

Read it later via Pocket or Instapaper.
Share it via Twitter or email.


In the Last 9 Months, Working With Leaders in the Stripe Developer Productivity Org, I Advocated for and Established a Team and a Measurement Framework to Understand and Improve the Software Engineering Experience (Thread)
5 minutes read.

Rebecca Murphey with a fascinating thread that made me go into the rabbit hole of reading Microsoft's "The SPACE of Developer Productivity." -- thinking of how to measure and improve productivity is incredibly hard to get right. The target can change to build a learning organization that delivers valuable and sustainable outcomes. As Rebecca mentions, measuring the SACE parts (without the P) might ease the adoption and start developing better tracking and learning frameworks.

Read it later via Pocket or Instapaper.
Share it via Twitter or email.


The Truth Curve and the Build Curve
5 minutes read.

Reading Giff Constable's post will get you to think about how you invest time in "building versus learning" and how much confidence you have in what you've just learned. This is very helpful to ensure we don't rush to build for no good reason while pushing beyond the analysis-paralysis. I like Phil Bergdorf's diagram, which captures it well.

Read it later via Pocket or Instapaper.
Share it via Twitter or email.

Jobs [sponsored]

Staff Engineer, APIs @ Forter
Our APIs process over $250B of the largest e-commerce merchants in the world every year, isolating fraudsters and abusers and protecting customers. Join us to help us build the Trust Platform for digital commerce.
 

 Looking to hire for your team? Promote your open positions on SWLW! 

Peopleware

Extreme Questions to Trigger New, Better Ideas
14 minutes read.

Jason Cohen shares helpful thought exercises to play with in your next brainstorming session. It (also) opens up a big "what if" when thinking of how to design a solution that is already proved a clear value and now needs a very different optimization factor, e.g., how can I make a system that is X10-X100 cheaper without hurting other dimensions is a significant way?

Read it later via Pocket or Instapaper.
Share it via Twitter or email.


Better/Sooner/Cheaper/More
3 minutes read.

"if you dictate time, budget, & scope then quality goes to hell & you lose control of time, budget, *&* scope. Instead, manage by managing scope." -- Kent Beck's comment on his post is an interesting one. Given that scope tends to expand over time, learning to play with it while controlling reasonable quality (to maintain and operate) and delivery time (customers waiting, other departments need to take it to market) is an important craft to improve at.

Read it later via Pocket or Instapaper.
Share it via Twitter or email.


Learning Starts With Desire
4 minutes read.

"Just as eating contrary to the inclination is injurious to the health, so study without desire spoils the memory, and it retains nothing it takes in." -- an inspirational post by Stew Fortier when we think of our learning. I see how it affects me. I also see how it affects my kids when I try to teach them something. This desire to learn different things shifts over time, so leverage it while it's there. Don't beat yourself up when the timing is off.

Read it later via Pocket or Instapaper.
Share it via Twitter or email.

And finally, inspiring tweets...

@noahkagan: Most college grads think WAY too much about their salary. Instead, they should optimize for environment. When you get a chance to work with smart people on interesting problems, TAKE IT. Your future self will thank you (and so will your bank account).

@JadeRubick: An executive is (or should be), a person who makes happen what needs to happen. This implies the skills of an executive are determining what needs to happen and knowing how to make it so. Seems simple, but it isn’t, at all.



p.s. if you're interested in joining SWLW's Slack channel, simply reply to this email and let me know. If you're leading a team, consider writing your Manager README (it's free) or getting my e-book and interviews Leading Snowflakes: The New Engineering Manager's Handbook. You can also support me by becoming a SWLW Patron. Thank you ❤️




Keep reading, keep learning.
-- Oren Ellenbogen.