What's happening

October 2013. Uber is hosting an offsite for employees in Miami. Travis Kalanick, then-CEO, sends a pre-offsite memo to attendees laying out behavioral guidelines for the trip. The memo is written in the casual register of a peer message, including hashtags and joke-asides. Nearly four years later, in the aftermath of the Susan Fowler post and the Holder Report investigation, the memo is leaked to Recode.

What this reveals

The memo's existence is more instructive than its contents. The contents are the symptom; the existence is the diagnosis.

When a CEO needs to send a memo telling adult employees not to have sex without explicit consent and not to sleep with direct reports, the question is not whether the memo is well-written. The question is what is happening inside the company that requires this memo to be written. There is an observation in organizational-health literature worth naming: the documents an organization produces about behavior are often inversely correlated with the behaviors it has institutionalized. Companies with strong cultures don't need to write down what good behavior looks like; companies with weak cultures write extensively about it. The memo exists because the underlying behavior could not be assumed.

The fact that Kalanick chose to write it in a casual register — #CEOlife #FML — is a separate signal. It tells you the CEO did not see the underlying behavior as the kind of serious cultural problem that requires a serious cultural document. The register-mismatch is what made the memo lethal in press coverage four years later. When Recode published it in 2017, the gap between the substance (preventing harassment) and the tone (peer-group joke) was the news. A memo with the same substance in serious register would have been forgettable. A memo with this substance in this register became one of the artifacts the Holder Report used to characterize the company.

Compare to Reed Hastings's Netflix culture deck, widely circulated in the same era. The deck addresses adult workplace behavior with sober structure, treating the topic as serious operating discipline. It became canonical because the register matched the stakes. Same delivery vehicle (CEO writing about behavior), opposite effect.

The transferable lesson

Two things, both larger than the surface lesson.

First, and largest: the memo you have to write is information about your company. If you're drafting a note telling adults not to do things adults should not need to be told not to do, the underlying problem is not the absence of the memo — it's the presence of the conditions that require it. Memo-writing as a substitute for cultural intervention is a category error. The memo creates a paper trail that, in litigation discovery, does more damage than silence would have, because it suggests the leader was aware of the problem and chose a low-effort response.

Second: the choice of who writes a culture document is more strategic than the choice of what the document says. Senior HR writing about behavior reads as policy and is forgettable in the right way. CEO writing about behavior reads as cultural signal and is unforgettable in the wrong way if the register is wrong. Leaders should write culture documents only when they can write them with the seriousness the topic requires; otherwise, the document belongs to HR or General Counsel.

There is a broader organizational design principle here. The companies that have weathered modern HR scandals well (Microsoft post-2014, Salesforce, Stripe) have institutional infrastructure that catches register-mismatches before they ship. The companies that have not (Uber pre-2017, WeWork, Theranos, Twitter pre-2022) treated CEO-as-culture-author as a feature rather than a risk. The generalization for any company past a hundred employees: the CEO should not be the primary author of culture documents. The CEO should be the primary enforcer of the behavior the documents describe.

What we don't know

The excerpts that survived are the ones Recode chose to publish, and they were chosen specifically because they were the most quotable. The full memo may have contained more serious sections that did not make the reporting. Even allowing for selection bias, the existence and tone of the quoted material are themselves the editorial substance. The lesson is about the craft of CEO writing on cultural topics, drawn from the artifact that became public — not about what Kalanick may have privately understood about the underlying conditions inside Uber.