CharlieKey takeaways from your favorite podcasts
← 20 October 2026Invest Like the Best

The New War for Silicon Valley’s Best People

Vanta CEO Christina argues that in the AI era software companies survive by shipping constantly, refusing to mark up tokens, and winning talent and trust with data, while valuations and hype distort decisions.

Watch on YouTube ↗

Charlie's summary

  1. You're only as good as the last thing you shipped: execution, integrations and a data foundation beat being first

    Christina says being first doesn't matter when someone can “vibe code your thing in a weekend.” Vanta kept winning by shipping constantly and by investing in integrations, now 400 tools feeding what it calls its trust graph, a security data lake. In retrospect, that data foundation let Vanta build its agent faster than competitors.

    ▶ 2:34
  2. Marking up tokens is not a business model: users will compare you to the model and you'll lose

    Christina says customers quickly say you've just marked up a Claude token 4x and go test the model themselves, which puts you in competition with the labs. Credits that obscure the exchange rate don't fool them. Vanta eats the token cost and takes lower margin. It has about 25% of customers in the agent flow and wants 75%, betting on better retention. Pricing can be sorted out later.

    ▶ 0:00
  3. Outcome pricing is what everyone wants, but nobody can agree on the unit of work

    Christina, once a microeconomics professor, loves paying for outcomes for the incentive alignment. The hard part is defining the unit: a document drafted, an alert closed, an account offboarded. Naming three means you could name 47 more, and users don't want that cognitive overhead. Her view is that charging for tokens puts you against Claude, so it is not clear how anyone gets there.

    ▶ 14:16
  4. Handing teams a safe option makes them build the old thing: tell them to burn the boats and bet on the LLM flow

    Vanta's team kept proposing onboarding wizards when asked for better onboarding. Christina found the blocker: they didn't think they were allowed to fail, and wizards were known to work. She told them to bet on the LLM flow. The result is a Claude-like canvas where customers drag in messy files and screenshots, and it works well. Find the real blocker rather than assuming a lack of motivation.

    ▶ 20:27

Give your email to unlock the rest of the key points — for every episode.

Do more with Charlie

Free — sign in with your email, no password.