Modern Tech Leadership — cover art: a 50-ish Rhialto (Steve's storybook-wizard self) in a deep cornflower-blue robe and hat, at a writing desk in warm lamplight, gently shelving a freshly-rolled scroll among a long row of similar scrolls; a single envelope lies crumpled in the wastepaper basket at his feet. 👍

2019 · Medium · Essay

“Email has devolved into what is essentially a poor man's feed for your other, better channels.”
— From Modern Tech Leadership, July 2019
Read the essay

© 2019 Steve Yegge. Originally published at Medium.

Author’s note

This is still a good read. Grab did some really smart things. The discovery we made around gentle escalation techniques was powerful.

Some things to be aware of. First, Go links have a commercial version now. Second, the Grab Mental Models (GMM) were eventually abandoned because people felt it was too much reading. And third, WPC really sucked, and we gave up on it after I posted this, and went back to Slack.

Otherwise, this was pretty solid. And I think GMMS could work at some places.

AI Notes

The promised follow-up to One Year Later: how OneGrab — a single distributed leadership team across Bangalore, Singapore, Beijing, Seattle and elsewhere — actually holds together day to day. The opening provocation is that millennials have stopped reading email, and that this is a useful data point rather than a generational complaint. Email has technical defects (upside-down threading, bolted-on attachments, no inline comments) and a social one — "the world's most finely-tuned machine for pissing people off." Grab's answer is four interlocking mechanisms. Workplace Chat (Facebook's Slack-equivalent) carries the conversation, with escalation groups that remove the us/them barrier. Grab Mental Models (GMMs) — Theo Vassilakis's invention — are 3-to-10-page documents in clear lifecycle states (IDEA, DISCUSS, REWORK, AGREED, ABANDONED); the line that travels is "GMMs act a lot like laws. It's OK not to read them or understand them, as long as you're not violating them." GSuites discipline is the third leg — Docs/Sheets/Slides with enforced structure, single ownership, explicit state, and a rule that long email is a doc. And go/ links, the searchable URL aliases Steve never stopped missing from Google, are the fourth.

The closing thesis: "An organization's velocity tends to be gated on how fast information can flow, and how fast teams can achieve alignment when they encounter friction." The post half-jokingly asks whether Grab should turn the GMM corpus into a leadership book.

Related listings

  • 2019

    Google to Grab: One Year Later

    The companion to this one, written two months earlier. One Year Later teases OneGrab as a leadership idea; Modern Tech Leadership is the promised follow-up explaining the actual mechanisms. Read in order for the full picture.

  • 2011

    Stevey's Google Platforms Rant

    The earlier essay where Steve argues that mechanisms are how culture compounds. GMMs are the Grab-era version of that thesis turned into practice — a writing-down system that lets a 24x7 org agree on things without holding meetings in every time zone.

  • 2018

    Why I Left Google to Join Grab

    The leap that made this dispatch possible. Read together, the three posts trace the arc — the bet, the look back, and finally the operational answer to how the bet actually runs day to day.