Sacred Cows as Your Roadmap
In a calcified org, the list of things nobody is allowed to touch IS the product roadmap.
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 6
- Confidence
- 95%
Kayvon Beykpour inherited a Twitter product org whose stated strategy was 'refine the core' and where users complained the product hadn't changed in eight years. Rather than starting from a blank-page roadmap, he inverted the problem: enumerate every assumption the organisation treats as untouchable, then deliberately schedule work against them one at a time. Each attempt doubles as a cultural X-ray, surfacing where hesitation, politics and career fear actually live.
Origin
Developed by Kayvon Beykpour during his tenure as Head of Product and GM of Consumer at Twitter (2018-2022), after joining via the Periscope acquisition. It produced hide replies, the ranked/reverse-chron toggle, Spaces, Communities, Community Notes and Super Follows.
Core principles
- 01A calcified org has already done your prioritisation for you — in reverse.
- 02Sacred cows are free roadmap: no user research needed to enumerate them.
- 03Attacking a sacred cow tests two things at once: the hypothesis, and the culture's tolerance for testing hypotheses.
- 04You cannot change culture from a pocket of the company; alignment must come from the top.
- 05The bet failing is acceptable; not being allowed to try is not.
- 06Keep refining the core in parallel — the goal is added ambition, not abandoned fundamentals.
How to run it
- 1
Enumerate the untouchables
Run the question explicitly with the team: what are all the things we believe we are not allowed to change? At Twitter the list included reverse-chronological timeline, 140 characters of text, users having zero control over replies to their own tweets, and Twitter never annotating a tweet.
Pro tip Write them down as a literal list. Ambiguous taboos are enforced socially; written ones can be argued with.
Watch out Do not confuse a sacred cow with a load-bearing constraint. Some things are untouchable for a real reason — force the reason to be spoken aloud, then judge it.
- 2
Secure top-level alignment before the first swing
Get the CEO/exec layer to explicitly sanction the campaign. Beykpour spent his first year 'walking through mud' because he was driving consensus across a functional org with no single tie-breaker.
Pro tip If you cannot get a tie-breaker, ask for a structural change instead (GM model) — that is the change that actually unblocked him.
Watch out Trying to change culture with only product-management reporting into you, without engineering or design, is 'changing culture with one hand tied behind your back'.
- 3
Start with a small, innocuous cow
Pick a low-stakes taboo first — his was 'hide replies', letting a user hide an unwanted reply on their own tweet. Small scope, but it directly violated a core orthodoxy, so it produces maximum cultural signal at minimum cost.
Pro tip Quick wins are the fuel: once people see a cow survive being touched, taking swings becomes addictive.
- 4
Read the resistance as data
Watch what surfaces. On hide replies, an engineer told the PM leading it that working on the feature was bad for her career and it would never launch. That reaction — not the feature metrics — was the real finding about the org.
Pro tip Escalate career-fear signals immediately and publicly. Silence ratifies them.
- 5
Sort people onto or off the wagon
Tell the story of the new ambition repeatedly, internally and externally. Then identify who is inspired by it and who is quietly pessimistic, and either convert them fast or move them out.
Pro tip Repetition is the mechanism: 'you can't just tell that story once.'
Watch out Beykpour rates himself and Twitter as 'terrible' at this step — they lacked both the org structure and the fortitude, which made cultural change far slower. High performers who are aligned will leave if you make them wait.
- 6
Run cows and core in one portfolio
Keep investing in the boring core (ranked timeline ML, notification relevance, onboarding, a broken SMS verification flow blocking signups in the UAE) while the sacred-cow bets run alongside. The aim is a balance, not a replacement.
Pro tip The core work funds the credibility that lets you take the big swings.
In the wild
The orthodoxy was that a user must never be able to hide anything said to them — only report it, with Twitter acting as policeman, which was unscalable. Beykpour's team took it on anyway. Weeks in, the PM leading it reported that a colleague had warned her the feature was career-damaging and would never ship.
→ The feature shipped and still exists in the product today, and the incident became the clearest microcosm of the cultural problem he was there to fix.
For years Twitter's stated product strategy was refine the core — turn the knobs that already work. It did return the company to DAU growth via the ranked timeline, but it also hardened the org's reticence to take any risky bet, and users said the product hadn't changed in eight years.
→ By working the sacred-cow list, the org shipped Spaces, Communities, Community Notes, Newsletters, Topics, Fleets, Super Follows and Twitter Blue — while continuing to grow the core.
Common mistakes
Treating a growth-positive strategy as proof that risk is bad
Refining the core genuinely worked at Twitter — and the fact it worked calcified the organisation's resistance to speculative bets even further. Success at optimisation is not evidence against ambition.
Trying to change culture from a pocket
Without top-level alignment and a decision-maker who will tie-break, a functional org turns every sacred cow into a consensus negotiation. That is politics, not execution.
Tolerating career-fear talk
The worst outcome is not a failed experiment — it is one person dissuading another from running the experiment. Left unaddressed, that single conversation prices in the whole culture.
Is it for you?
Best for
A newly-appointed product or functional leader inheriting a large, risk-averse org that ships incremental optimisations and is publicly criticised for standing still
Not ideal for
Early-stage startups with no orthodoxy yet, or orgs whose fundamentals are genuinely broken and need the core fixed before any ambition is credible
From the transcript
“there were so many sacred cows at Twitter that the sacred cows are like their own road map it's like a built-in free road map”
“sacred cows and you know the the process of starting to address those one by one reveals a lot of the cultural hesitations that that…”
“you can't change culture without having alignment from the from the top”
From the episode
Twitter’s former Head of Product opens up: being fired, meeting Elon, changing stagnant culture, building consumer product, more
Kayvon Beykpour