Go All-In When the Experiment Works
Big companies experiment plenty — they fail by hedging instead of doubling down on what works.
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 4
- Confidence
- 90%
A transformation playbook that corrects the real failure mode of large companies. The myth is that big companies don't experiment; the truth is they experiment a lot but keep hedging when an experiment works instead of going all-in. The framework: kill the pocket veto, tie every employee's personal success to adopting the change, and be deliberately top-down on the one bet that matters.
Origin
Jeetu Patel's account of making Cisco (90,000 employees) an AI-first company, treating ChatGPT's November 2022 launch as the seminal go-all-in moment.
Core principles
- 01Innovation is a binary choice you make every hour, independent of company size.
- 02Large companies experiment plenty; the failure is refusing to double down when an experiment works.
- 03A 'pocket veto' kills conviction — in a big company, ask enough people and someone always says no.
- 04Align personal success with the change: adopting it protects your job, ignoring it makes you irrelevant.
- 05Be clear on what is up for debate and what is not.
- 06On the one existential bet, deliberate top-down clarity beats consensus.
How to run it
- 1
Separate what's debatable from what's not
Declare clearly which decisions are open for debate and which are settled, so conviction bets don't die by pocket veto.
Pro tip In a large company, if you ask enough people, someone will always say no — so protect high-conviction bets from death-by-committee.
Watch out Leaving an existential bet 'open for debate' lets it get quietly vetoed.
- 2
When an experiment works, go all-in
Once a bet is validated, stop hedging and double down fully rather than keeping options open.
Pro tip Cisco didn't hedge on AI — they said 'we're going all in,' using ChatGPT's Nov 2022 launch as the seminal moment.
Watch out The classic large-company failure is experimenting a lot but never going all-in on what works.
- 3
Align personal and company success
Reassure people the change won't take their jobs, while making clear that not adopting it will make their role irrelevant.
Pro tip Frame it as guaranteed: get dextrous with the new capability and you're safe; don't, and your job won't stay relevant long-run.
Watch out If people fear the change threatens them, they'll quietly resist; you must remove the fear and set the incentive simultaneously.
- 4
Be deliberately top-down on the one bet
Even if you dislike hierarchy, get the entire company literally on the same page that this is the direction — no ambiguity.
Pro tip Patel says he deep-down doesn't respect hierarchy, but chose to be very, very deliberate and top-down specifically on the AI-first bet.
Watch out Reflexively avoiding top-down clarity on the one existential change leaves the org fragmented and hedging.
In the wild
Rather than hedge, Patel got all 90,000 employees on the same page that Cisco is an AI-first company, tied AI dexterity to job relevance, and treated ChatGPT's November 2022 launch as the trigger to go all-in.
→ Cisco transformed from an older, slower, traditional enterprise into an AI-forward company — with a product about to ship 100% written with AI.
Common mistakes
Hedging after an experiment succeeds
Large companies run plenty of experiments but keep their options open when one works, so the winning bet never gets the concentrated commitment needed to pay off.
Letting fear drive silent resistance
If employees believe the change threatens their jobs, they quietly resist; you must remove the fear and make clear that non-adoption — not the change itself — is the real career risk.
Is it for you?
Best for
Leaders driving a major technology or cultural transformation through a large, established organization.
Not ideal for
Genuinely uncertain early experiments that haven't yet produced a clear signal — going all-in prematurely is just unhedged gambling.
From the transcript
“innovation in my mind is a choice... Every day you come into work and you can choose to be thinking about being creative or you…”
“What large companies don't do is when an experiment works they don't go all in and double down. They try to keep hedging. We didn't…”
“if you didn't use AI, if you weren't going to be dextrous in whatever job function you're doing, then your job is probably not going…”
“you can always have a pocket veto in a large company where if you ask enough number of people people say no.”
From the episode
AI is critical for humanity’s survival: Cisco president on the AI revolution
Jeetu Patel