Real Alpha
Watch modeSubscribed

Cheeky Pint · 4 min · Oct 20, 2026

Mercedes stops engineering work that misses the lap-time target

Toto Wolff

Team Principal & Co-Owner, Mercedes-AMG PETRONAS F1 Team

Led Mercedes to eight consecutive constructors' titles from 2014 to 2021; part-owner and team principal since 2013.

  • John CollisonHost · Co-founder & president, Stripe

    Sold his first startup, Auctomatic, at 17, then co-founded Stripe in 2010.

About this segment

Wolff says projects must earn resources through simulated lap-time gains, then stay correlated with results. Teams stop work that serves a department rather than the car.

From “How Toto Wolff built the most valuable team in F1”, Oct 20, 2026. Watch the full episode

“It needs to fundamentally contribute to the overall target of lab time rather than what we discussed before to the department's target.”
Toto Wolffat 24:10 · Cheeky Pint24:10

Takeaways

  1. Estimate lap-time value in simulation before building.
  2. Recheck predictions against wind-tunnel and track results.
  3. Stop experiments that miss the team's overall target.
Watch 4:13 from 21:03

In this episode

Full episode
  1. 11:59Mercedes tested a clever loophole before putting it on the car4:59
  2. 28:59A CFO’s warning led Wolff to exit his dot-com investments5:18
  3. 46:27F1 teams combine scarce-asset appeal with strong profits6:52
  4. 0:12F1 spots driver talent through pace, feel and teamwork4:25
  5. 6:15Mercedes screens engineers for values as well as ability2:22

  6. 21:03Mercedes stops engineering work that misses the lap-time target4:13

  7. 37:55Ownership gives Wolff room to run Mercedes entrepreneurially5:00Up next
  8. 1:08:02Wolff expects two or three people to share his role3:10

Transcript

Transcribed by machine (Whisper) from the episode’s audio, so a word may be off here and there.

  1. 21:07

    But the other risk is just technical risks. You know, will it work? And I think in any R&D organization, you know, we have this at Stripe where we only have so much time, we only have so much money, we have only so much engineers, and we need to put them on the very, you know, towards the projects that are most likely to work. And you don't know up front what's going to work. And so part of the special sauce is that ability to predict this product direction is a good one versus this one is just, I mean, it's not awful.

  2. 21:36

    It's just not the best. And you're fundamentally playing this game of where you spread the poker chips. How do you manage technical risk? How do you think about allocating those very precious chips? And again, they're even more precious now in the new Cascapula. So, the unfortunate part of it is that your business is far more complex, in my opinion. Why? Because you deal with the end customer or whoever the customer is, an organization, an entity, or a person.

  3. 22:06

    Because you can define technical specifications that you believe make the process better or more efficient or successful. But there is someone at the end that actually needs to buy your product. and that person or that organization is going to measure whether the humans that are actually his customers or her customers are going to like it. How can you predict that part? With us, it's all data and the simulation. It's much more measurable. It's much more measurable.

  4. 22:32

    That at a very early stage, we will define what is the lag time contribution that we imagine of that part. And that all starts in the virtual world. It's all simulation, whether it is CFD or whether it's the wind tunnel in the way we measure it or we attribute downforce and drag or whether it is mathematical models that we have. And all along the process, you continue to correlate with your expectations.

  5. 23:01

    And there are rare occasions where you believe that something eventually when it ends up in the car is going to be a tenth of a second in average overlap quicker and you realize it's not. or you realize that the driver doesn't feel comfortable with it and isn't feeling, you know, isn't feeling safe enough to drive it. So we need to consider also if the end customer, which we know exactly, you maybe not, will be able to extract from that virtual, let's say theoretical lap time benefit, but it's much more measurable.

  6. 23:41

    Engineers like to fall in love with their ideas and I just need one more simulation run, or I just need one more, you know. We don't have that. But do you have to kill people's ideas because they're in love with them, but they're just not going to work, or you just don't have the resources to staff it? I think we have people like this who, if they are also forceful, they might... I'd put the organization under pressure to actually continue with the experiment beyond a reasonable point.

  7. 24:06

    But they are not part of the organization anymore. This is one of the personality traits that we hold dear. It needs to fundamentally contribute to the overall target of lab time rather than what we discussed before to the department's target. so there will be a decision within the group and then in the wider technical management to say we're stopping this now so if i was to talk to an engineer at mercedes i would find them surprisingly dispassionate and analytical where they're not falling in love with a particular technical approach or they're not falling in love with their idea it's either it works or it doesn't exactly you fall in love with the result yeah it's that these extreme emotions and gratifications that you feel when you win a race or a championship.

  8. 24:57

    And therefore, the kind of rational part of the engineers, I would hope, is superior against, let's say, an emotional falling in love with your development part. I want to ask a bit about your investing background