Process vs Chaos In Startups

Michael Seibel 0:00 A Google style of carefulness was being applied by a startup whose sole job was to explore the risky paths.

Dalton Caldwell 0:09 You can end up where the process becomes this big, and the actual product is this big. It, it'd be like if at-

Michael Seibel 0:15 Yes

Dalton Caldwell 0:15 ... at the bolt factory, there was 1,000 process people-

Michael Seibel 0:18 Yes

Dalton Caldwell 0:19 ... and one guy making bolts.

Michael Seibel 0:20 And one bolt a day comes out of the factory. [laughs]

Dalton Caldwell 0:26 [laughs]

Michael Seibel 0:26 [upbeat music]

Michael Seibel 0:26 Welcome to Dalton + Michael. Today, we're gonna talk about, we might even debate-

Dalton Caldwell 0:31 Ooh

Michael Seibel 0:31 ... slightly-

Dalton Caldwell 0:32 [laughs]

Michael Seibel 0:33 ... process, procedure.

Dalton Caldwell 0:34 The P word.

Michael Seibel 0:35 Bureaucracy.

Dalton Caldwell 0:36 Yep.

Michael Seibel 0:37 Meetings. And the kinda, the prompt for this one is, um, confusing an organized process with an effective process, and confusing chaos with ineffective or bad outcomes.

Dalton Caldwell 0:52 Yeah, I mean, I think the context here is you and I have debated this for many years-

Michael Seibel 0:57 Yeah

Dalton Caldwell 0:57 ... as colleagues.

Michael Seibel 0:58 Would people guess which s- I don't know that people-

Dalton Caldwell 1:00 Yeah, what do you think? Who's the pro-process and anti-process person?

Michael Seibel 1:02 Yes, and who's the chaos guy?

Dalton Caldwell 1:04 Yeah, who's... I think they-

Michael Seibel 1:05 I think they would assume I'm the chaos guy. I think they might.

Dalton Caldwell 1:08 People who know me would never. [laughs]

Michael Seibel 1:09 It's a great point. You're right. I actually have no idea if people understand-

Dalton Caldwell 1:14 Yeah

Michael Seibel 1:14 ... that I'm the anti-process guy. [laughs]

Dalton Caldwell 1:16 [laughs]

Michael Seibel 1:16 And that you're the pro-process guy. I don't know if that's obvious. Beats me, man.

Dalton Caldwell 1:20 Tell us in the comments. Like [laughs] tell us in the comments.

Michael Seibel 1:24 Um, and so yeah, so the general debate is Michael will be like, "Oh, we need a structured process to do X, Y, and Z."

Dalton Caldwell 1:31 Yes.

Michael Seibel 1:31 And I'm like, "Why?"

Dalton Caldwell 1:32 Yes.

Michael Seibel 1:32 "That sounds bad."

Dalton Caldwell 1:34 We would end in many different places.

Michael Seibel 1:35 Yeah. [laughs]

Dalton Caldwell 1:36 [laughs] Depending on the case.

Michael Seibel 1:36 There'd be some process and then no process.

Dalton Caldwell 1:38 Yeah.

Michael Seibel 1:38 Yeah.

Dalton Caldwell 1:39 But it was always an interesting debate because it was always a, I always like the debate because process should always be treated skeptically.

Michael Seibel 1:46 Yeah. It should always-

Dalton Caldwell 1:48 Otherwise, it'll inf- otherwise it'll-

Michael Seibel 1:50 Exactly

Dalton Caldwell 1:50 ... take over.

Michael Seibel 1:51 Someone has to be the devil's advocate-

Dalton Caldwell 1:52 Yep

Michael Seibel 1:52 ... for any new process, and you played that role excellently. [laughs]

Dalton Caldwell 1:57 [laughs] You know. Fair enough. Fair enough.

Michael Seibel 2:00 So how is people thinking about useful process versus bad process?

Dalton Caldwell 2:03 Useful process, so to take the opp- so to, to argue the opposite side that I'm-

Michael Seibel 2:06 Yes

Dalton Caldwell 2:06 ... usually on, useful process is the following: once a best practice exists-

Michael Seibel 2:12 Yes

Dalton Caldwell 2:12 ... you should just do the best practice.

Michael Seibel 2:14 Yes.

Dalton Caldwell 2:14 Otherwise, you're just wasting a ton of time. Think about, um, automation, like, uh, the Industrial Revolution-

Michael Seibel 2:22 Yeah

Dalton Caldwell 2:22 ... where if there's one way to make a bolt on an assembly line-

Michael Seibel 2:24 Yeah

Dalton Caldwell 2:24 ... just keep making the bolts.

Michael Seibel 2:26 Yeah.

Dalton Caldwell 2:26 Those bolts are gonna be better-

Michael Seibel 2:27 Just do it exactly. Yeah

Dalton Caldwell 2:27 ... in every way than a bespoke welder trying to make a bolt by hand. And so process in its best form is just establishing a state-of-the-art best practice-

Michael Seibel 2:38 Yes

Dalton Caldwell 2:38 ... and ramping it way up, and there's no question about how to do something properly.

Michael Seibel 2:45 Well, like, process in its best form, I love that bolt example, produces a great bolt-

Dalton Caldwell 2:51 Which is better-

Michael Seibel 2:52 ... every time. [laughs]

Dalton Caldwell 2:52 Yeah, it's, it's perfect.

Michael Seibel 2:54 Yeah.

Dalton Caldwell 2:54 And, and the process identifies, uh, errors.

Michael Seibel 2:57 Yes.

Dalton Caldwell 2:57 Like, part of the process of making the bolt is the camera detects, oh, this bolt's bad, and it throws it off the assembly line. So yeah, the, an assembly line is an example of fantastic process.

Michael Seibel 3:07 And a, a great test for whether you have a good process is the outcome good [laughs] -

Dalton Caldwell 3:11 Yep

Michael Seibel 3:11 ... and, and getting better.

Dalton Caldwell 3:12 And consistent.

Michael Seibel 3:12 Yeah, and consistent. [laughs]

Dalton Caldwell 3:14 Yeah. [laughs] And you, and you know it's provably good.

Michael Seibel 3:15 Yes.

Dalton Caldwell 3:16 Like, you can actually measure that it's effective.

Michael Seibel 3:18 Yeah. What about bad process?

Dalton Caldwell 3:20 Bad process is just like any classic bureaucracy where it exists to grow itself. It's almost like a, dare I say, cancer-

Michael Seibel 3:29 Okay

Dalton Caldwell 3:30 ... where it's a machine that all it wants is to have more power or more people in the process, and it becomes a vector to basically take over everything such that, you know, someone has to weigh in on every decision and has control over how decisions are made.

Michael Seibel 3:47 Yes.

Dalton Caldwell 3:47 There's a whole lot of people employed-

Michael Seibel 3:50 Yes

Dalton Caldwell 3:50 ... in trying to keep the process going. It's like a jobs program. And so you can end up where the process becomes this big, and the actual product is this big. It, it'd be like if at-

Michael Seibel 4:01 Yes

Dalton Caldwell 4:01 ... at the bolt factory, there was 1,000 process people-

Michael Seibel 4:05 Yes

Dalton Caldwell 4:05 ... and one guy making bolts.

Michael Seibel 4:06 And one bolt a day comes out of the factory. [laughs]

Dalton Caldwell 4:08 [laughs]

Michael Seibel 4:08 Yes.

Dalton Caldwell 4:09 That would be bad process.

Michael Seibel 4:10 I think an int- a, an interesting thing about process is when people reach for it as a solution. And I can say from personal experience, there are a couple different times where I reach for it as a solution. Um, so one is, like, the same problem keeps happening. The second is fear or anxiety. If I'm doing something for the first time and I'm afraid that it's not going to work, sometimes process gives me comfort. Well, have I made a plan? Like, you know, if I'm cooking something for the first time, have I looked up recipes, and da, da, da, da. Um, so personal fear and anxiety, but I also think fear and anxiety for other people.

Dalton Caldwell 4:52 Mm.

Michael Seibel 4:52 Like, I actually think process can reduce anxiety. That doesn't necessarily mean the output will be good. [laughs] But I do think that, um, a lot of process comes from people being afraid and anxious-

Dalton Caldwell 5:08 Yep

Michael Seibel 5:08 ... and wanting the comfort. And chaos might be comforting when the result's good. It's rarely comforting during the chaos. [laughs] It, it's often, especially for a certain type of people, it's often indi- anxiety-inducing during the chaos.

Dalton Caldwell 5:26 Yeah.

Michael Seibel 5:26 Um, on the flip side, though, what's so funny is if you were to ask me to think back through the biggest innovations or things that have triggered the biggest innovations or change or positive things, they've all started with chaos. They've all started with something breaking-

Dalton Caldwell 5:44 Yeah

Michael Seibel 5:44 ... and having to not follow-

Dalton Caldwell 5:47 Yeah, my, my-

Michael Seibel 5:47 ... whatever the plan was. [laughs]

Dalton Caldwell 5:48 My mental model for this is the more something looks like producing bolts-

Michael Seibel 5:53 Mm

Dalton Caldwell 5:53 ... the more process works.

Michael Seibel 5:55 Mm-hmm.

Dalton Caldwell 5:55 And the more that it looks like producing art-

Michael Seibel 5:58 Mm

Dalton Caldwell 5:59 ... or a hit song-

Michael Seibel 6:01 Yeah, yeah, yeah

Dalton Caldwell 6:02 ... the less you can-Be like, "Here's our-

Michael Seibel 6:05 Process

Dalton Caldwell 6:05 ... here's our process." And the- and again, if you think about what people are mad about with, you know, the entertainment industry, Hollywood, or the music industry, it's process-

Michael Seibel 6:12 Yes

Dalton Caldwell 6:12 ... seeping its way.

Michael Seibel 6:13 MCU, yes.

Dalton Caldwell 6:14 Right? [laughs]

Michael Seibel 6:15 The eighth Avengers movie, yes.

Dalton Caldwell 6:16 That's like, that's process-

Michael Seibel 6:18 Yes

Dalton Caldwell 6:18 ... screwing up art. And so that's the continuum, is how much is this art and how much is this a bolt?

Michael Seibel 6:24 I like that, 'cause if I- if I go back to my new favorite term, MBAification, there is this kind of theory that any output could be optimized with process. And like yeah, if you can make a hit movie with that process, I feel to make a better movie with-

Dalton Caldwell 6:41 Yeah

Michael Seibel 6:41 ... with great process.

Dalton Caldwell 6:42 Yeah, let's just license some, some comic book guys and-

Michael Seibel 6:46 Let's get Brad Pitt. [laughs] Let's, you know, like-

Dalton Caldwell 6:47 And that's literally the entertainment industry these days.

Michael Seibel 6:49 [laughs]

Dalton Caldwell 6:49 It's, it's like, like MBAification of creating art.

Michael Seibel 6:52 Yes. And so I think that it's tricky because, um, to your point, it's funny 'cause when you're saying art, I might say innovation. Process makes it kind of hard to invent a new way of doing things. When you need to invent a new way of doing things, you kind of need the freedom to maybe do stuff that's unpopular.

Dalton Caldwell 7:09 Yep.

Michael Seibel 7:10 That is scary, that may be more expensive than people... Right? Like, all the things that process tried to control for. [laughs]

Dalton Caldwell 7:16 Yeah.

Michael Seibel 7:17 And so I think what gets tricky is like, okay, sometimes I need to be using my creative brain, sometimes I need to be using my bolt brain. How do I choose, right? And it's like well, you know, when you're trying to do your, you know, fundraising paperwork, I'd probably be using my bolt brain, right? Like-

Dalton Caldwell 7:34 Yes. [laughs]

Michael Seibel 7:36 [laughs] It's like you're gonna have to write up your own legal docs from first principles. But like if I'm trying to make my customer service go better, am I innovating? Or sh-

Dalton Caldwell 7:46 I mean, you, you definitely wanna measure it like it's bolts.

Michael Seibel 7:48 Yeah.

Dalton Caldwell 7:49 Like, I think for a lot of process you wanna start with measurement.

Michael Seibel 7:52 I agree. But like do you just wanna fire the customer service guy from the other company? It's not a bad place to start.

Dalton Caldwell 7:59 It's tricky though.

Michael Seibel 8:00 Yeah, yeah, I understand. It's tricky. Like, I always kind of will ask a founder where are the top three places you wanna innovate? Like, like, you probably can't innovate across 20 zones. If you think customer service is gonna be your differentiator, maybe you don't hire that 20 year-

Dalton Caldwell 8:18 Yep

Michael Seibel 8:18 ... person. But if it ain't [laughs]

Dalton Caldwell 8:22 [laughs]

Michael Seibel 8:23 Yeah. Like, you know, w- where do you wanna put your innovation points? I also think that there's something about process, as a lover of process, I love iterating process. Like, one of the things I really enjoy is, okay, the factory is producing bolts too slow.

Dalton Caldwell 8:42 Yeah.

Michael Seibel 8:43 How do we tweak the factory? And I think that sometimes when I encounter people who don't like process, they don't engage in iterating the process. And so it's a tricky thing, 'cause like some of these things have to be done bolty.

Dalton Caldwell 8:59 Yep.

Michael Seibel 9:00 And if you're not tweaking it and fixing it, so now you've got a bad factory. You know, it's like we know a lot of people who are like, "No meetings ever." And then it's like, "And no written specs 'cause I don't wanna be held [laughs] admin-" It's like-

Dalton Caldwell 9:13 I mean, this may be t- too in the weeds, but I think a great example of good process is, um, code reviews.

Michael Seibel 9:20 Yeah. [laughs]

Dalton Caldwell 9:21 [laughs] A written spec. Like, w- no, no, let's not even go to the spec part. But basically, you know, the, the default thing a startup does is everyone can just ship to production, and then you break the website, and then you're like, "Yeah, maybe not that." And you, you have to do a couple of layers before you can ship to production.

Michael Seibel 9:37 I might even argue just staging.

Dalton Caldwell 9:39 Yeah. [laughs]

Michael Seibel 9:40 But any- but, but I think this is a, you know, this is an interesting point where it's kind of like being too extreme on either side screws you. Like, I think what was so fun working together is that we could play, like, process dude and, like, chaos dude-

Dalton Caldwell 9:54 Yeah

Michael Seibel 9:54 ... because we knew the middle would be in, like-

Dalton Caldwell 9:57 Yeah

Michael Seibel 9:57 ... we, we would kind of get the best of both worlds. I think that both of us without the other one would've had to be a little bit more in the middle. [laughs]

Dalton Caldwell 10:06 I think that's well said.

Michael Seibel 10:09 [laughs]

Dalton Caldwell 10:10 Um, and, and where this comes into play when you think about it for a lot of folks out there is, uh, innovator's dilemma, which is a lot of businesses that are gonna get killed by AI disruptors, they're gonna have to ignore all the process people, right? This is cl- you know, to, to recap innovator's dilemma for everybody, it's if a business is really successful doing A, and then this new B thing comes out and it seems potentially disruptive-

Michael Seibel 10:35 Yeah

Dalton Caldwell 10:35 ... businesses will tend not to do B even though they know they should because it screws up A. Like, there's too many stakeholders that are excited about doing A, and that's, the whole business is A.

Michael Seibel 10:45 Yes.

Dalton Caldwell 10:45 So why would we blow up our own business?

Michael Seibel 10:48 It opens the door, right? It opens the armor in the bus- It's funny because that can happen for startups too. I was talking to a startup in the kind of ed tech space and, you know, AI is enabling them to do things that inconceivable two years ago. But they were saying to me like, "Oh, well, the kids really wanna f- they really wanna read Pokemon books, but of course we can't show them Pokemon 'cause of the copyright." And I'm like, "I respect Pokemon. I respect copyright, for the record."

Dalton Caldwell 11:17 Okay.

Michael Seibel 11:18 [laughs]

Dalton Caldwell 11:18 All right. I duly noted, Michael Seibel.

Michael Seibel 11:21 [laughs]

Dalton Caldwell 11:21 It's on the record.

Michael Seibel 11:22 I think my record is muddy on this topic. [laughs] But, like, it was interesting that, like, a, a Google style of carefulness was being applied by a startup whose sole job was to explore the risky paths.

Dalton Caldwell 11:39 Yeah.

Michael Seibel 11:40 And I think, like, th- this is a kind of meta point that I keep on telling founders is like, "Hey, the, the, the beautifully blazed, straight, like, Autobahn-style road, the big company gets to walk that road. You are chopping down things, and you're in the for- you're blazing the path. And, like, you might have to bend some rules to blaze that path. But if you're walking the Autobahn path, they're gonna-

Dalton Caldwell 12:05 Yep

Michael Seibel 12:06 ... they're gonna run you over every time. Like, you represent the risk in the economy. [laughs] Like, that is your job. You're the risk. Embrace it. Um, 'cause when you don't embrace it, you're just lying to yourself. You still are the risk, even when you're walking the Autobahn. [laughs]

Dalton Caldwell 12:21 Yeah. [laughs] Amen.

Michael Seibel 12:23 So, um, it was an interesting thing. I think I convinced him to let p- the kids read Pokemon books.

Dalton Caldwell 12:30 [laughs]

Michael Seibel 12:31 [laughs] All right. That sounds like a good wrap-up. Thank you, Dalton.

Dalton Caldwell 12:36 Sounds good. Thanks, Michael. [outro music]