Progressive Delivery, with Heidi Waterhouse
Episode 120 · February 18th, 2026 · 1 hr 8 mins
About this Episode
Whitney and Coté talk with Heidi Waterhouse, co-author of the book Progressive Delivery.
You can watch the video of the recording as well, if you're into that kind of thing.
Heidi on the World Wide Web:
LinkedIn.
Her Website.
The book, Progressive Delivery.
Transcript
Whitney Lee 00:00
Cote, are you, um, like on your personal cell phone, are you like a Mac person or like Android?
Coté 00:07
Oh, IU I've, I've embraced the Apple lifestyle.
Whitney Lee 00:10
Yeah. Yeah, I have too. It's kind of like my family has, and if I don't, then I turn the chat green and that, you know, upsets everyone anyway, that
Coté 00:17
green.
Whitney Lee 00:19
And so, uh, that was smart of them to work in the social pressure of conformance because Uhhuh, it's working, but that's not the point. The point is, recently on my phone I used the Apple, and recently on my phone I had to do an upgrade. The upgrade introduced this liquid glass, which is like a new UI uhhuh. If you could see my face right now, I'm squinting and disgust. Um, it really upset me. Did you, have you upgraded to Liquid Glass?
Read Full Transcript Hide Full Transcript
Coté 00:49
Yes. Yes. I, I, I, I guess the embrace part was, I just, I just, you know, I'm, I'm all in, just as soon as possible. I upgrade all the stuff and I just go for it.
Whitney Lee 00:58
And tell me about that mindset.
Coté 01:01
Uh, well, first, a part of that mindset, maybe to go to the second question is like, I know there's a lot of people who don't like it, and there's lots of complaining, and I feel like, I feel like even if you're way deep into the Apple world, like maybe some VP got fired over it or something. All sorts of speculation.
Whitney Lee 01:18
Good.
Coté 01:18
Uh, but if you were to ask me now what Liquid Glass is, I couldn't tell you. I have no idea. Like, like the, the only frustrating, and I'll get back to your question, like I did encounter a frustrating thing earlier today on the iPad where. The keyboard keeps getting small and it's really hard to get it bigger. And so maybe they could fire someone about that. Or, you know what? They don't have to fire them, they could just fix it. That's, that's all. But, uh, yeah, so about, about upgrading like all the time, I just, I do get excited about like the new features that you get and I also know that there's often undocumented little features.
So it's like, I don't know, like, like maybe here's an analogy and then I'll stop. Like you remember Toys or Us, it's a giant toy store, and I remember this stopped happening at some point in the late eighties, but I remember you could go in there and it was like, Costco is for adults. There was, mm-hmm. There was just like so many things and they had rotating inventory. And you could go around there and just find things that you weren't expecting. And so, uh, it's kind of a little bit of that when you upgrade. It's like, oh, look at this setting. You don't know what, you don't, what framing you down. And you know,
Whitney Lee 02:32
you don't know what's available to you until it's in front of you and you like that.
Coté 02:35
Yeah.
Whitney Lee 02:35
Choice. Yeah.
Coté 02:36
And, and often these things are very lame. I don't, I don't know. I just upgrade the stuff.
Whitney Lee 02:41
I just, yeah. I am so resistant to upgrading. I've, uh, like for liquid glass, as far as I know, it's just a change in how things look. I don't know that there's actual any change in functionality, but I'm a visual person. I have an art degree, like I care a lot how things look and I put a lot of time and care and I'm making it look the way's true. I like it. That's true. Yeah. And then this liquid glass came on over and just like blew everything up. And my time is valuable and I don't wanna sort through the menus and get it back to how I had it.
Yeah. But also looking at it makes me angry. So I don't know which thing I'm gonna do, which path I'm gonna take moving forward.
Coté 03:17
Yeah.
Whitney Lee 03:18
And a little bit I've given into, like back in the day when upgrades were rare and it was more of a thing that you could manage, I did try to manage it because I, I wanted that control and it, this is something that's a big piece of my life and honestly, I've kind of given up and been like, you like just freaking upgrade. I can't even think about it. Hopefully it doesn't blow anything up. And Liquid Glass is a recent case where it has, but often something is upgraded and I just don't have the cog the bandwidth anymore to keep track of what it is.
Coté 03:47
Yeah.
Whitney Lee 03:47
Um, uh, which I think the world in general is doing to us. That's,
Coté 03:51
that's probably my thinking is like, you know, this is, uh, it's more entertainment to be upset about this than anything.
Whitney Lee 03:58
It's gonna get upgraded on me anyway. Eventually fine. Just do it, but I don't like it.
Coté 04:03
Yeah.
Whitney Lee 04:04
Um, uh, our guest today. Hello Heidi. Will you introduce yourself and weigh in on our conversation please?
Hedi Waterhouse 04:13
Yes. So my name is Heidi Waterhouse and I am a dev tools marketer advocate, and I am the co-author of a new book called Progressive Delivery, which is about this very problem and how much, um, we don't consider the user when we make the software. And, and it's, it's really interesting to me what you're saying, Whitney, because one of my volunteer gigs is, um, sex education,
Whitney Lee 04:44
uhhuh,
Hedi Waterhouse 04:44
like comprehensive sex education. And the thing that we keep saying over and over again to these young people is consent, consent, consent, consent, ongoing enthusiastic consent. Yes. And I have not ever. Felt like that was actually a thing that I got from software. I get coerced consent. I'm like, do you wanna use the software? Well, I guess I'm collecting your data now. I'm like,
Whitney Lee 05:12
yes,
Hedi Waterhouse 05:13
that's not cool. Or,
Whitney Lee 05:15
or do you wanna read these terms and conditions that are, you know, to move forward and using something that's become a vital part of your life? Or read these five pages of legal technical documents? And even if you disagree, then what? Yes.
Hedi Waterhouse 05:26
Right.
Whitney Lee 05:27
I, yes.
Hedi Waterhouse 05:29
Yeah.
Whitney Lee 05:29
Um, yeah. Uh, I've always, hmm. I never really thought I could have control. What does that look and feel like?
Hedi Waterhouse 05:38
Right? So I think starting with the liquid glass thing, which boy, you are not the only angry person out there. I was, I was just at a, um, art camp, which was amazing by the way, uh, with a bunch of retired age people who had been in technology. And they were so mad because the way the, um, apple, uh, phone camera operates had changed somewhat.
Whitney Lee 06:09
Mm-hmm.
Hedi Waterhouse 06:09
And they wanted to take pictures of their quilts and they couldn't figure out how to do it. And they're helping each other. And I'm like, I can't help you. I'm, I'm an Android person, but I, I was like, the pressures inside software are very much like, what are we releasing this quarter? And how am I getting promoted about it?
Whitney Lee 06:30
Yes.
Hedi Waterhouse 06:30
And it's so disconnected from, I'm trying to do a thing in my daily life. Why did you move my fricking cheese?
Coté 06:39
Right. Right.
Whitney Lee 06:40
Yes.
Hedi Waterhouse 06:40
Like, this is not a new problem.
Whitney Lee 06:43
Yes.
Coté 06:44
Yeah. And someone, someone had a phrase re I forget who it was, they had a phrase, uh, this is a couple weeks ago, of like backseat driver software. It's like constantly like asking you things like, would you like to leave a review? Would you like to do this? Have you checked this out? Here's a tip. And just like always getting in your face about like itself, like in, in, in my, like, the worst is, uh, Uber Eats. It's constantly sending notifications. 'cause if I turn off notifications about Uber Eats, I'm not gonna get my Thai food. Right.
So like, I've gotta have this like avenue in there.
Hedi Waterhouse 07:19
I
Whitney Lee 07:19
just,
Hedi Waterhouse 07:19
I keep them turned off. They used that channel.
Whitney Lee 07:21
That's right. Exactly. Mm-hmm.
Hedi Waterhouse 07:23
Like you gave them consent for telling you when your Thai food is arriving, but you did not give them consent to, did you know that if you recommend five people, like No, I don't want that.
Whitney Lee 07:34
Here's my question though. Once something is, is. At the center of your life, like your cell phone, like what motivates them to get user consent? Like they would have to do something so bad, and maybe they did with liquid glass to like get someone to move their entire life onto a different system. So at some point they, they have an advantage that they know they have and they're taking advantage of.
Mm-hmm. As a consumer, how do you fight that and, and how does that work into your, your idea that consent is? What's the consequence of not getting that consent? I guess
Hedi Waterhouse 08:10
so I think that's why, why we need to talk about it to kids because like they don't, they don't live in a world where consent matters. Like, I don't wanna do that. Yeah. I'm sorry kid. You are having a shower. Like you smell that. Mm-hmm.
Whitney Lee 08:25
Yeah.
Hedi Waterhouse 08:26
Um, and so, and, and the surveillance culture that we are subjecting children to right now Yeah. Is priming them to believe they don't have any right to consent. Mm-hmm.
Coté 08:39
It's
Hedi Waterhouse 08:39
like, I know where you are, I know what you're doing. I can see what you're doing. They don't have any privacy. And it is heartbreaking when I talk to my young adults about how much they assume they're being surveilled, how little privacy they think they have and they're not wrong.
Whitney Lee 09:00
Yeah.
Hedi Waterhouse 09:01
But. I'm that cusp generation where you could be anonymous on the internet and you could like go for a 10 mile bike ride without your parents knowing where you were. And
Whitney Lee 09:13
Uhhuh.
Hedi Waterhouse 09:16
So like to bring it back to software instead of culture. I think, um, when I first got Android phones, when I first started getting phones, I would jailbreak them every time because I, I had a very specific keyboard I wanted to use.
Whitney Lee 09:31
Okay.
Hedi Waterhouse 09:31
And I'm like, you don't give me access to my keyboard. I'm jailbreaking the system. Uhhuh. And eventually. They got to the point where they're like, okay, this is an app. You can have your keyboard without having to jailbreak. And I stopped jailbreaking and I, I didn't, I wouldn't know how to break Jailbreak and Android anymore.
Whitney Lee 09:49
Yeah.
Hedi Waterhouse 09:49
Like it used to be, you know, I would get a new phone and new phone day was jailbreaking it and putting all the stuff I wanted on. I don't do that anymore because they are allowing me to have the experience that I want.
Coté 10:03
Yes.
Hedi Waterhouse 10:03
That I can turn off location tracking, like I live in occupied Minneapolis Uhhuh. The way I use my phone has really changed because I know that there is unusual and dangerous surveillance going on.
Whitney Lee 10:18
Mm-hmm. Like,
Hedi Waterhouse 10:21
do I take my phone, which is, we all think of it as sort of a safety device. Like if something happens to me, I can call for help. Do I take my phone to stand guarded the food shelf? Do I take my phone to distribute groceries? No, because the very fact of my phone going places that distribute groceries means that somebody can see where a person who needs groceries is, which implies that they are not going out.
Yeah.
Whitney Lee 10:49
Which
Hedi Waterhouse 10:49
implies that they are not here feeling safe.
Whitney Lee 10:53
Yeah.
Hedi Waterhouse 10:55
And so that consent thing is still possible, but it's very all or nothing. I can't just turn off
Whitney Lee 11:04
You
Hedi Waterhouse 11:04
can't location tracking because it's still pinging a tower.
Whitney Lee 11:09
And so what happens is people do, at some point, there's some sort of lever where you abandon the technology altogether if you're not able to consent in the ways that you want to consent. Um, yeah,
Hedi Waterhouse 11:22
like,
Whitney Lee 11:22
so
Hedi Waterhouse 11:23
the, I'm currently researching mesh networks, like how much, yeah. How much could we get off big internet? Like I never thought about that last year.
Whitney Lee 11:31
Yeah. And what's upsetting too is, is data can be pulled from years past when we weren't worried about it or upset about it or protecting ourselves in that way. Yeah.
Hedi Waterhouse 11:44
So from a software creator point of view, I think it's really interesting to say the more changeability, the more configurability we give people, not just at a interface level, but at a data level
Whitney Lee 12:00
mm-hmm.
Hedi Waterhouse 12:00
The less likely we are to have them give us up entirely. Like the less likely people are to table flip and leave our environment.
Whitney Lee 12:09
Mm-hmm. And,
Hedi Waterhouse 12:10
and,
Whitney Lee 12:10
and maybe there's, oh, sorry.
Hedi Waterhouse 12:14
Yeah. Right now that's, um, that's like only very like politically intense. People care about this.
Whitney Lee 12:22
Mm-hmm.
Hedi Waterhouse 12:23
But we'll get to the point where the quilt grandmas are like, okay, I can trust Android's camera more than Apple's camera, and I don't care if I turn my chat green because. This is what needs to happen.
Whitney Lee 12:38
Yeah. I can trust this thing I've learned isn't going to change on me, or these settings I've chosen aren't going to come out from underneath. And I would, I would postulate that the more like heavily adopted the center of your life, a technology is probably the more room you have to try to manipulate your users before they table flip. If it's not something you use often, if it's not what you want, pretty immediately you, you table flip much faster than if it's something at the center of your life. But the, the idea of consent being important stands in either use case and the whole,
Hedi Waterhouse 13:12
I think Facebook is a very interesting example of this.
Whitney Lee 13:15
Oh, I've already, I've table flipped Facebook. I don't know what's going on over there anymore.
Hedi Waterhouse 13:19
And I think, I think that's an interest. But did you do Instagram?
Whitney Lee 13:24
Uh, I've table flipped Instagram a couple years ago, but I did do it for a while. Yes.
Hedi Waterhouse 13:27
Yeah. And I, I had, I never signed into the Facebook system, like that was
Whitney Lee 13:33
okay.
Hedi Waterhouse 13:34
I, whatever. That was my nerd. Um, but now I'm on WhatsApp because that's how my neighbors communicate.
Whitney Lee 13:43
Ah. And I'm like,
Hedi Waterhouse 13:44
oh, you got me back in because of the network effect.
Whitney Lee 13:48
Uh, Uhhuh what?
Hedi Waterhouse 13:49
So I think,
Whitney Lee 13:50
Hmm. And what do you mean by the network effect?
Hedi Waterhouse 13:52
I mean, if I want to communicate with my hyperlocal, um, response group, it's on WhatsApp because it's got better multi-language support than signal.
Whitney Lee 14:04
Okay. Mm-hmm.
Hedi Waterhouse 14:07
And I'm like, okay, this
Whitney Lee 14:08
is that. And so now you're in the me back in the meta ecosystem. Yeah.
Hedi Waterhouse 14:13
And I hate it, but the, the rewards are higher than the aversion.
Whitney Lee 14:21
Yeah.
Hedi Waterhouse 14:21
But see, I think it's interesting, Cote, did you have a Facebook,
Coté 14:26
uh, ever? Yeah, I've got a Facebook thing. Yeah. I I never log into it anymore. It happens by accident. Right. And so sometimes I go to like Instagram and I, I always think like something's gonna happen here that I care about, and then it doesn't, so then I leave. I don't know. But yeah, I've got that stuff. And of course I live, you know, I live, I live in Europe, so I have WhatsApp. Mm-hmm. The universal communication protocol of not America, of Europe, basically.
Hedi Waterhouse 14:55
Yes. So I think that we can see that Facebook tweaked their algorithms and denied consent so long that even Normies are starting to leave. We're not normies. Yeah. But like my, my mom abandoned Facebook.
Whitney Lee 15:14
Wow.
Hedi Waterhouse 15:14
I'm like, okay. You have, you have screwed up your core demographic now.
Whitney Lee 15:20
Yeah. What was the straw that broke the camel's back in terms of your mother or what's bothering that generation?
Hedi Waterhouse 15:28
Yeah. Um, the push for engagement bait over family news.
Whitney Lee 15:34
Mm-hmm.
Coté 15:34
Right, right.
Whitney Lee 15:35
Yeah.
Hedi Waterhouse 15:36
Like absolutely, like the reason people were using it was family community connection that doesn't pay as much as rage bait.
Whitney Lee 15:46
How does this relate? Oh, sorry.
Hedi Waterhouse 15:48
And we know that Facebook is pushing rage bait because it pays more like they've said that. Mm-hmm.
Whitney Lee 15:55
So, so how does this idea of consent based computing relate to in acidification? It seems like almost the same thing.
Hedi Waterhouse 16:03
Um, it is the consumer side of, in acidification, because when Doc Row talks about in acidification, he's talking about like. It was valuable. And then the economic forces of honestly, venture capital are forcing this hypergrowth model, Uhhuh that rewards anti-social in sort of the technical term. Like we want software that is prosocial for the most part, Uhhuh. Like we want connection, we want to talk to humans, we want to feel affirmed.
We wanna know what's going on at our hometown. Right? These are all prosocial desires. Mm-hmm. And it isn't just social media, but it's largely social media that we think of. This way,
Whitney Lee 16:54
Uhhuh
Hedi Waterhouse 16:55
antisocial behavior drives higher engagement and turns off your reasoning so that you are more likely to be like, I'm going to buy 50 res,
Whitney Lee 17:10
yeah. I'm gonna give this up more attention instead of real humans in my life. More attention.
Hedi Waterhouse 17:15
Right. But you feel like you're going to it for real humans, and then you get sucked into this reactionary state. And it's really good for software like Uhhuh. It's, it's good for selling ads, which is the only way you can really do hypergrowth.
Whitney Lee 17:32
Yeah.
Hedi Waterhouse 17:33
If, and, and Dr says this in, in ification, um, but if we were trying to do moderate growth the way that business and capital worked for most of the history, we wouldn't have this problem because it would self-correct, but because there's all this VC money coming in, we don't have that signal that says people don't like this because VCs like it.
Whitney Lee 17:59
Yeah. Vc. What VCs think matters more than what users think.
Hedi Waterhouse 18:04
Exactly.
Whitney Lee 18:05
Wow. Okay. '
Hedi Waterhouse 18:05
cause you're not getting as much money from the user as you are from the vc.
Whitney Lee 18:09
Yeah.
Hedi Waterhouse 18:10
So the signal is very skewed.
Whitney Lee 18:15
But then eventually it needs to convert from a VC funded product to a user funded product. How does that switch happen?
Hedi Waterhouse 18:22
The VCs don't care 'cause they get their money out before that point.
Whitney Lee 18:26
Wow. Right. The world is bananas.
Hedi Waterhouse 18:30
It's, but like, if you think about it, um, everybody who did early Uber Uhhuh made their money back hundreds of times, right?
Whitney Lee 18:41
Mm-hmm.
Hedi Waterhouse 18:43
And they may still be a little invested in it, the investors, but
Whitney Lee 18:46
Yeah. Yeah. That's,
Hedi Waterhouse 18:48
but on the whole, they're onto the, the next thing in, the next thing, they're out, they got their, their hypergrowth money, and now the company is left without that, you know, sustaining money.
Whitney Lee 19:04
Mm-hmm.
Hedi Waterhouse 19:05
And they have to figure out how to make money. And the only way to make it is from exploitation of the users.
Whitney Lee 19:12
Yeah. And, and then we're back around to user consent and ification. Yeah.
Hedi Waterhouse 19:20
Yep.
Whitney Lee 19:22
But
Coté 19:22
we still, we still need a heuristic for It's okay to change the ui. Yes. As, as a, as a big stable company that, uh, you know, it, it, I I don't know if you would say like, apple has been in ified. It's, it's in some other class of, of that Apple is weird, weird
Hedi Waterhouse 19:39
case. Yeah.
Coté 19:40
And, and so It does, it does. But, but we gotta solve the, the Whitney problem. Like, is she gonna leave? Maybe, maybe we could call this to Whitney line one day she'll be like, they upgraded, uh, iOS and I'm out. Mm-hmm. And it's just, what, what, what would that line have to be like? And that's very
Hedi Waterhouse 19:57
important for
Coté 19:58
them to know.
Hedi Waterhouse 19:59
I can tell you how to avoid the line, which is to say. Whitney hates this ui, there's a button that will flip it back. Mm-hmm. And there's
Whitney Lee 20:09
mm-hmm.
Hedi Waterhouse 20:09
No technical reason for, especially for a UI change that you couldn't have both versions
Whitney Lee 20:17
at, at least for a while. But isn't there a point where you're maintaining two versions and, and you don't wanna do the work involved in keeping that old version around?
Hedi Waterhouse 20:27
I don't know. Do you like having those users
Whitney Lee 20:31
too?
Hedi Waterhouse 20:31
Okay. So there, there is an argument for Sun. Um, but I think that on the whole, we are so averse to repeating ourselves that we software makers move so much faster than software users.
Whitney Lee 20:49
Mm-hmm.
Hedi Waterhouse 20:50
That we completely lose sight of what it's like to only buy a phone every force.
Whitney Lee 20:57
Yeah.
Hedi Waterhouse 20:59
That's like the world average is more like six, but the US average is like three.
Whitney Lee 21:07
Okay.
Hedi Waterhouse 21:08
So tech people who get a new phone every year are way out on the cutting edge
Whitney Lee 21:13
uhhuh.
Hedi Waterhouse 21:15
It doesn't feel like that when you're on the inside.
Whitney Lee 21:18
Yeah.
Hedi Waterhouse 21:19
It feels like, oh, everybody is doing this now. Everybody I know has a phone with 5G. Why wouldn't we do this?
Whitney Lee 21:30
Yeah. So there's a real bias in the, all the mindsets of people working in tech because of the people that they interact with in the world.
Hedi Waterhouse 21:37
I always say that, um, the thing that I want to do to graphic to to UX designers, Uhhuh, is give, uh, like the oldest shittiest monitors currently available. Like, welcome to your VGA Hell, because how users are experiencing the world.
Whitney Lee 21:56
Absolutely. Oh yeah, that's so wise. And, and there is some sort of testing involved where you test on the oldest version of whatever, but like, that's like an automated test of um, it's past fail, fail situation's. Automated test. Yeah.
Hedi Waterhouse 22:09
And it's not the developer, it's not the designer that's doing that. Designers have the giant screens with like the curve super high def. Right? And they're, they're live in a different world than people who are like, I actually, um, you know, this is on a roll cart in a hospital, or this is in a car mechanic shop, or, you know,
Whitney Lee 22:37
so I wanna get around to how do you incorporate the user's feedback into your software delivery lifecycle. But I think maybe before we do that, let's introduce your book that you've, uh, co-written recently. We talk about the progressive delivery book.
Hedi Waterhouse 22:53
Yes. So the Progressive delivery book is the evolution of, uh, me and Adam Ziman and James Governor and Kim Harrison. And we noticed that there was this thing happening where we didn't need to do big releases. We were doing sort of agile releases, but we weren't pulling back any of that information into the software development lifecycle. So like, if you think about the DevOps loop, right? Mm-hmm. There's, there's dev and there's ops, and it goes around and it's like, and you develop, you deploy, you launch, you iterate.
Whitney Lee 23:35
Mm-hmm.
Hedi Waterhouse 23:35
That loop is missing one of the loops. There's a user user acceptance loop. So it should be like,
Whitney Lee 23:43
ah,
Hedi Waterhouse 23:43
you develop. You launch, you, you know, get it out in the world and then the user experiences it and feedback comes back into the ops loop and back into the devs loop uhhuh, because we can do that. We can absolutely tell if, say if there was a switch that said turn off liquid glass uhhuh, 50% of people are turning off liquid glass the minute they see it, that signal needs to get back to the depths.
Mm-hmm. Or if, um, people are having bandwidth problems like Uhhuh, you've done something to your game and you've put in a very, um, high bandwidth ad mm-hmm. That needs to come back to ops. And if ops is not instrumented and looking for that, that's a user problem. It's not a launching problem. Like it, the deployment worked fine.
Whitney Lee 24:41
Yeah.
Hedi Waterhouse 24:42
But. But the way that people are experiencing it is not working fine, and we could instrument and be watching for that. So
Whitney Lee 24:50
the,
Hedi Waterhouse 24:50
the and the
Whitney Lee 24:51
liquid glass example. Mm-hmm. Then when you're doing like a canary rollout and you're only pulling a couple percent of your users, that's when you can see people hate liquid glass as opposed to waiting for general release in the model that you're describing. Roll Yeah. Rollout. Yes.
Hedi Waterhouse 25:08
So the progressive delivery book has a bunch of concepts that are like, okay, in order to build a team that can listen to the user, here are the things you need. Abundance. Autonomy, alignment and automation. If you have a team that has those things mm-hmm. Then you can be listening to the user and, and the user experience beyond just what the user tells you. Because like, please do not ask me to tell you how I feel about my software. Like, now I'm annoyed. Mm-hmm. You've interrupted my flow. But you can't say like, you read all the way down the page, or you only read the top third of the page. Mm-hmm. Or you used AI to summarize that page. Those are really useful data points that you don't need to interrupt somebody to see.
Whitney Lee 25:56
Yeah, so, so this framework with the four a's that you just said, is that like, is that something like the Dora, where it's like you actually have specific things you measure on? Or is it more of a mental model for how to move forward? Describe that framework to me a little more. In practice,
Hedi Waterhouse 26:12
it's more of a mental model. We're working on trying to build an instrument to measure it, but it's difficult because every team is individual and honestly, like the Dora level measurement stuff is really sophisticated.
Whitney Lee 26:30
Okay. Yeah.
Hedi Waterhouse 26:31
But what we wanna say is like there are some heuristics that you can use, like how much money can you spend without asking about it.
Whitney Lee 26:42
So that's autonomy. Yeah,
Hedi Waterhouse 26:44
that's autonomy and abundance. Well, and it's also abundance. It's like, do you have the things that you need to do what you want? Like do you have to go through a tedious process to provision?
Whitney Lee 26:57
Okay.
Hedi Waterhouse 26:58
Right. And there should probably be a little check on the process, especially in the land of AI, where you can run up a real good, some real good bills before you realize it. Um, but like, can you do what you want to do without having to slow down and ask for a ton of permission? Uhhuh, do you know, do you know what that is? Like? That's the other thing is like the unlimited vacation problem. Unlimited vacation is such a scam. I'm so mad about it because really it just means we don't wanna have to pay out vacation time.
Whitney Lee 27:31
Yeah. Mm-hmm.
Hedi Waterhouse 27:34
Unlimited vacation means very different things in different teams and different organizations. And sometimes it means you can never take that much time off. And sometimes it really does mean unlimited, take what you need, but they look the same on paper.
Coté 27:49
Mm-hmm.
Hedi Waterhouse 27:49
So abundance is like, how do you know what the guardrails are so that you can get right up against them? Mm-hmm. So that you can be high performing without feeling like you're going to pop out of the looses track.
Whitney Lee 28:02
Absolutely.
Hedi Waterhouse 28:04
And then we get to like autonomy, which is do you have the power to change things? Do you have as a team or as a person, the power to say, this isn't working, or I feel like we could do better. And then alignment
Whitney Lee 28:23
it ends blameless.
Hedi Waterhouse 28:25
Yeah, exactly.
Whitney Lee 28:25
Be able to, not just the power, but to without consequence. Yes.
Hedi Waterhouse 28:29
Yeah. And alignment is about both alignment within the team. Like do we know what we're trying to do? Mm-hmm. And alignment with the user. Hmm. So we understand what the user is trying to do. Are we being considerate? Are we asking consent? So alignment is really, there's almost two alignments. There's like internal and external.
Whitney Lee 28:53
Mm-hmm.
Hedi Waterhouse 28:54
And you have to do different things to cultivate them. And then automation is, once you know which, how much of it can you automate so that you can get a flywheel?
Coté 29:03
Mm-hmm.
Hedi Waterhouse 29:04
When we were designing this, several of us were working with teams who were doing an agile DevOps transformation. There's a lot of companies out there who are not agile. Or, or they're doing some other kind of thing. So we wanted something where people could start at a, a small level, start on your own team and scale it up without necessarily needing to do a big transformation thing. It's like, okay, can you increase the autonomy?
Can you increase the alignment? Okay, having done that, can you automate what you've increased? And so we want people to have a bunch of different avenues to approach being more progressive and adding feature flags, and adding deployment monitoring and adding instrumentation for user experience. All of these things are going to help you. None of them is a transformation because,
Whitney Lee 30:05
yeah.
Hedi Waterhouse 30:06
Those projects are a nightmare. Anyone who's ever been on a transformation project is like, what if I just walk into the ocean instead? How's that sound?
Whitney Lee 30:17
And then, uh, and you're, it sounds like you're saying too, like any, any improvement you can make in any one of these areas is an overall improvement. You don't need to approve across all four altogether to e
Hedi Waterhouse 30:28
Exactly.
Whitney Lee 30:29
To improve. And
Hedi Waterhouse 30:31
we actually think that trying to do all of them at once is overwhelming and too difficult. Mm-hmm. But if you, if you step up, if you ratchet, then you can get a lot of improvement in relatively little pain.
Whitney Lee 30:47
And then does your book help define like a goal on the horizon that you're working toward? And what does that look like exactly?
Hedi Waterhouse 30:55
So the goal on the horizon is that your delivery process is. Simpatico with the acceptance process of the user. Uh, one of the books that we read when we were writing it was Future Shock Uhhuh, which is from, I don't know that book. No, because it's from 1969.
Whitney Lee 31:15
Oh,
Coté 31:18
it was, it, it was a available in mass quantities in use bookstores in the eighties. Just, uh,
Hedi Waterhouse 31:23
exactly. And, um, nothing
Whitney Lee 31:26
like reading a book from the past about the future, but it's probably still the past.
Hedi Waterhouse 31:30
We're not wearing paper clothes, but like some of the stuff he said about education was eerie. I'm like,
Whitney Lee 31:36
yeah.
Hedi Waterhouse 31:36
Oh, we tried that. We found out it didn't work, but we did try that. And Huh?
Whitney Lee 31:44
What, what specifically?
Hedi Waterhouse 31:45
Um, so he said that education was going to stop being so centralized and splinter into little, um, basically, um, pods.
Whitney Lee 31:58
Huh,
Coté 31:59
that sounds
Hedi Waterhouse 31:59
weird. And that every child would have individualized, uh, education. Oh. And it turns out that while that seems idealistic and awesome, you have to have a foundation of shared learning. And if you don't slow the bright kids down and speed the slow kids up, um, you have real, like problems extending out into further education. So
Whitney Lee 32:28
interesting. So, okay,
Hedi Waterhouse 32:29
so where does this come back? We're also not living underwater, but there was a ton of stuff about, he's like, people need options. People need the ability to control their life. And if you don't give them that, they will rebel and table flip
Whitney Lee 32:47
uhhuh, but do, okay, here's a, so I. Oh, I, when I'm home, I never use DoorDash, but when I travel, sometimes I do, and especially in the morning. Um, I get up at like four or five in the morning and, and I, I want coffee and it's not really available in the hotel yet, so I might DoorDash coffee, so I'm like door dashing Starbucks recently when I was in New York, and the number of options I have for every little thing is freaking wild.
It's, it's like, like so many milk options. But then do you want it steamed? And then do you want these syrups? And then do you want like half calf or full calf or like, like there were literally for like a cup of coffee. I like plain coffee. Like there were 50 options. Mm-hmm. Like, is that seems like overboard or do you think that's like a good thing?
Hedi Waterhouse 33:36
I think you want to go through that once and then you want Starbucks to know who you are when you log in.
Whitney Lee 33:43
Ah. That did happen. So I, user experience we wanted,
Coté 33:47
if you were using like, what is it, F-F-M-P-G on the command line, and you finally get it set up and then you just make an alias or a script file so you never have to touch it again, or Uhhuh or that YouTube, DLP, whatever, all all that, all that kind of stuff. You've have infinite options, but you sort of want like that default setting. My favorites
Whitney Lee 34:05
and Yeah, I, I get that in a software example, in a Starbucks example, when there's um, human, human like effort that needs to happen every time a coffee is ordered to like, to remember all of these different options and how it's doing, it seems like it could slow down the whole process.
Hedi Waterhouse 34:23
Yeah.
Whitney Lee 34:23
But in software, yes.
Hedi Waterhouse 34:24
The goal is to have persistent identity and persistent preferences.
Whitney Lee 34:34
Okay. Yeah.
Hedi Waterhouse 34:34
So I think of this as an accessibility thing. Um,
Whitney Lee 34:37
yeah,
Hedi Waterhouse 34:37
there are dark mode people, there are light mode people. There are people who don't care. I am. I'm a switches at Sunset person. Okay. Love it. Which many people find weird. They're like, how? And I'm like,
Coté 34:50
I was gonna say, I've always wondered who actually uses that. So, so, so confusing. Yeah.
Hedi Waterhouse 34:57
For one thing, it like notifies me that I should pay attention to the time I also have like
Coté 35:02
Yeah,
Hedi Waterhouse 35:03
that's
Coté 35:03
good.
Hedi Waterhouse 35:04
Decorative lights over here that switch every couple hours so that I am aware of the passage of time.
Whitney Lee 35:11
I love that.
Hedi Waterhouse 35:12
Yeah. But, uh, so in an accessibility sense, what I want is for all of my apps to know that about me and for them to also be like, yes, you want your font a little bigger.
Whitney Lee 35:26
Mm-hmm.
Hedi Waterhouse 35:27
And I want that identity to travel with me. I want there to be a universal usability language that says, uh, this person. Can't do high contrast because they get migraines or this person can only do high contrast because they have low vision uhhuh. Those are conflicting accessibility requirements. But why can't they be persistent? Like, why can't they be tied to who I am?
Coté 35:55
Right. It seems like a, uh, almost a set it and forget it type of like, uh, phrase you would use for it, right? Like you figure it out and you set it just sort of like, uh, with your coffee, Whitney, you, uh, you figure out the way that you want it and then like, you know, this is the one key word to use or like. The, the, the, the one thing that you wanna do. And then I guess you could also, like, I always try to have my small contained rebelliousness against the world order, and I always say I want a small instead of a tall.
And I think, you know, it makes me feel a little more empowered when it comes to engaging with the big squid beast about everything.
Hedi Waterhouse 36:30
Yeah. But so I think that the end goal of software should be that it fits to our hands, whatever our hands are, like it should fit to our hands instead of us contorting ourselves to fit to the software. And the way we do that is give people optionality, ask them for consent, and give them a lot of understanding of why people have things set the way they are.
Whitney Lee 36:58
Mm-hmm. Like,
Hedi Waterhouse 37:01
like every, um, IDE. People can figure it the way they want it, like Kote was saying. Mm-hmm. Like, you just, you have a thing that you want, you want to set it that way, but sometimes you learn from other people's. You're like, oh, I didn't know you could do that. And that
Coté 37:20
kind of like a little bit of vanilla syrup makes the coffee taste better. Just a tiny bit
Hedi Waterhouse 37:25
like, like quarter shot. Yeah.
Whitney Lee 37:27
There's a button for that tiny bit too,
Coté 37:31
you know, on, on, on this topic to kind of generalize it seems like, how, how do you account for the, the, I don't know if it's cyclical 'cause I don't know the time period, but the kind of like frequent forgetting of the user, right? Like, like we had, we had one example of the ification thing of like, we haven't so much forgotten about the user as. We need to make money, and now we're doing that, right? Mm-hmm. And so like, and in fact you think a tremendous amount about the user because you're trying to figure out how to like extract money from their presence and your involvement.
But so we'll put that over there as a negative case. But you know, a lot of things you've gotten over are like, uh, like we used to have u we still do, but there used to be a lot more conversation about user experience or design, especially when to go back to them, especially when Apple went through its ascendancy in the late two thousands. It was like, oh, design is important, right? Like maybe. And so it seems like, and then, and then I was thinking this when you were going over, like the DevOps loop is like, you're right. There was like, I don't remember early, early DevOps, but that user feedback loop didn't really ever become part of it, right?
And so I don't know like what accounts for that like. We as a technology community never go that last mile. Like we, we sort of like, are constantly like, remember the user, we should get back to that. And it just seems to get dropped out of most cycles that, that, that we have,
Hedi Waterhouse 39:04
the system does what it's designed to do.
Coté 39:08
Sure. Yeah. So, and, and, and so, so, so maybe like, so, uh, are, are the people designing it not paying attention or do they just not care Or like the designers, like what, what's up with them?
Hedi Waterhouse 39:22
The designers are under a lot of pressure to look as cool as the other cutting edge software people. And uh, when their boss comes in and says, Hey, can we do this? Or, Hey, you're going to do this. It's about comparing it themselves to other software and not to the user experience.
Coté 39:47
Right, right, right.
Hedi Waterhouse 39:48
Mm-hmm.
Coté 39:49
Because that's, there's, there's sort of, there's sort of like a negative cycle of like your, I forget, I forget the phrasing around this, but like you are, uh, you're doing your craft for other craftspeople, not for the mm-hmm. The users of it. Sort of like, if, if I was in the market for like, uh, well, I mean, I could make a direct analogy to software, but if I was in the market for like a bookcase, like I don't really need a fussy bookcase, I just need to put my book somewhere. But you could see that other bookcase people would be like, Ooh, you use the beveled knobs.
Tricky. That's hard to pull off accurately. Whereas, so I may, maybe that's the thing is if, if you're internally facing, you should be aware that
Hedi Waterhouse 40:26
Right.
Coté 40:27
Optimizing for your internally facing stuff is not that great. Which I think, you know, I mean to general, to, to put it in business terms, it's sort of like a lot for, you know, to kind of pick on them some of the Dora metrics. Like it's, it can be difficult to say. And how did this, like make the company more money? Or whatever, whatever like, uh, goal you might have as far as an organization operating,
Hedi Waterhouse 40:52
it's always more money, but yeah. All right. So I think of it this way. Nobody likes Confluence.
Whitney Lee 41:00
Mm-hmm.
Hedi Waterhouse 41:00
Confluence is the least bad solution that we have currently for the problem space. That is shared corporate knowledge.
Whitney Lee 41:10
Mm-hmm.
Hedi Waterhouse 41:11
It's better than SharePoint, usability wise. That doesn't mean it's good, but it is like the lowest common denominator of usability. And if you customize it, if you make it too far from like Confluence standard, then you have a lot of onboarding and difficulty getting people to understand what you're doing with it. Mm-hmm. Like when you bring people in, they're like, oh, I know how to use Confluence. Okay. You don't know how to use Confluence here.
Whitney Lee 41:41
Yeah.
Hedi Waterhouse 41:43
So. Confluence suffers from its own success in that they have added everything that anyone has ever asked for. It's not opinionated. It is 50 kinds of things to set your coffee. What do we do about the fact that like, there's, there's people who try by making a different, um, internal communications tool. Okay. But you run into the problem that the company itself is formed around a confluence like system.
Whitney Lee 42:19
Mm-hmm.
Hedi Waterhouse 42:21
And so I think there's a lot of the difficulty of changing things is that change is a variance and variance is more expensive than standardization.
Whitney Lee 42:34
Yes.
Hedi Waterhouse 42:34
So you have to choose really carefully what variance are a business value and what is just like rebelliousness. Like what adds business value? And currently we all use Confluence because it's good enough and nothing has really come along. That's better. But we will change when something comes along that's better. Like we changed off of SharePoint.
Coté 43:06
Do do you think, do you think that says something about the in, in software, the value of the software category? That to say we're fine with, I, I'm gonna say we're fine just as a way that of summarizing a lot of what you're saying, but to say that we're fine with something we collectively don't like, like does it indicate that, like, that category of software isn't that valuable? Or, or does it just mean that someone hasn't come up with something better?
Whitney Lee 43:32
It's, it's ripe for dis disruption.
Hedi Waterhouse 43:34
It's, it's ripe for disruption, but disruption is very hard because there's a lot of mass. Like, there's so many, there's decades of influence or decades of information in there that we would have to transfer out,
Coté 43:52
right? Right. So you, you'd have to, you'd have to move the data, but then you also have like, just the, in addition to that, you'd have the switching cost of, uh, skill of using it, right? Mm-hmm. And, and familiarity and things like that. And, and so I guess I don't know all the terms for this, but, but it, it feels like, I guess the analysis is like, how much effort would it take us to be better?
Right. And, and like, like, would that
Hedi Waterhouse 44:16
pay?
Coté 44:17
Right, exactly. And, and, and, and if that amount of effort, if we don't wanna do it, then we don't, we don't value this stuff enough to put that effort in to get to better, like, we're fine with, with what we have. It's kinda like, I guess to go outta software. You have people who are, uh, car people and not car people, and they're like, I, I need a, uh, a fancy BMW or I could have a Ford focus. And you know, maybe there's people who are like Ford Focus Premier vehicle, you don't know what you're talking about, but like it does, it does seem like you're pretty much fine with the Ford Focus. You're kind of saying this category of cars not important. Like there or not as important as like 500 other things in my life and I don't worry about this one so much.
And so like, it would seem like maybe when it comes to like internal knowledge management for technical people we're kind of like, eh, confluence is fine. If this was a more important area of our lives, we would put in the effort to make it better. But really whatever.
Hedi Waterhouse 45:18
Right? Like think about how many ides people cycle through because that's very important to them in their daily
Coté 45:26
work, like core to what they do. So any small amount of uh, gain you get whatever gain may mean just happiness or productivity mm-hmm. Is probably worth the effort of switching or experimenting with switching it to it.
Hedi Waterhouse 45:38
Right. But also. It's an individual choice, right? You're not necessarily constrained to use this tool or that tool when you're writing your code, right?
Whitney Lee 45:48
Right. Right. Mm-hmm.
Hedi Waterhouse 45:50
But you do have to share a knowledge management system. So you have to have consensus.
Whitney Lee 45:59
So to get back to the premise of the book, um, DevOps loop is missing a loop or missing a section where there's not user feedback being collected, and then user feedback being fed back to the people making the software, which I very much agree with. And actually, I, I beg that drum a little bit in the context of platform engineering. Mm-hmm. But you're, you're big nodding Yes. Um, but what I wanna understand is how does adding that loop and getting that user feedback back to developers affect the bottom line?
Hedi Waterhouse 46:34
So. This is a theory like we do not have the tool to prove this, but making the wrong thing is very expensive and the faster, this is a thing we learned from Dora, the faster we release in smaller iterations, the more information we have about how to make the right thing and we don't go down the wrong path. And so when we get that user feedback and feed it back into the developer, we can see that people aren't adopting or that they are adopting in much higher numbers than we expected. And then we can lean into what's working and pull back from what's not working or figure out why it's not working. I think that the other thing that I would love to see and I never do see is developers talking to people who use their software and anger.
Whitney Lee 47:27
Yes.
Hedi Waterhouse 47:27
Um. You like, there's the salespeople and you know, and you're selling to an it uh, manager, but they're not the person who's like actually doing the work. And the example I have from this is from a long time ago, but it really stuck with me, which is I worked for a company that was making a skin on top of the Medicare system here in the us. And it was, it was pretty, and it was mouse driven because what's underneath that is green screen, 10 key, like that is what runs Medicare. And they're like, oh, this is terrible. We can, we can make this better. We can put a, a friendlier, more useful predictive interface on it. What they hadn't done was talk to the people who actually did medical coding.
Whitney Lee 48:20
Oh, they don't want pick ops.
Hedi Waterhouse 48:23
No, they're, they're wizards of the 10 key. And they were like, mm-hmm. Why would I, I don't even use my mouse. What's a mouse? Um, but they were so fast at what they did that any kind of like gooey was slowing them down. And I'm like, yeah, you just built a whole software product around the misconception that modernity is automatically better.
Coté 48:49
So how, how, how might one, let's say like progressively change that, right? Like let, let, let's assume the bar to entry to being really good at the green screen, tin screen, 10 key thing, it takes some time to learn, right? Like you could, yeah, you could analogously say like, well, the mainframes work, right? Like if, if, you know, if, I guess they do, if you know, if you know what you're doing, this is great. Mm-hmm. And, and, but, but then you get another problem in the system of like, but if we introduce someone new into the system. It's difficult for them, and they would like to swipe instead of click. I don't know.
I'm making up stupid phrases here. Right. But like, like what would be the way of sort of like, uh, shifting between those two things? Do you just have different interfaces or like, how do you, I think you have, how, how do you introduce liquid glass to Whitney without her realizing it's happening?
Hedi Waterhouse 49:39
I, I think you make it more gradual. So I think that you have like, here's the, the all mouse way to do this, but did you know that there are keyboard shortcuts? You keep doing this mouse action. Did you know there's a keyboard shortcut for that? Would you like me to put a reminder in the corner for you for that shortcut until it gets into your fingers?
Coté 49:59
Yeah. Yeah.
Hedi Waterhouse 50:00
Okay. Here's the next one. Here's the next keyboard shortcut. So that's a lot of user education. How we solve the Whitney problem is to say, first of all, like, would you like to phase into this? Okay. Like in a year, everything's going to be liquid glass. Would you like to, um, change your icons right now, but not everything else? Mm-hmm. Or would you like to, you know, so that the one month it's okay and the next month it's okay. And you're learning at a rate of change that you find acceptable. Mm-hmm. Because this was a big part of our thesis when we wrote the book, is that it's not that people are averse to change. It's that, I mean, they are, but it's not just that.
It's that the rate of change is too high and we are being asked to adapt too quickly. And our, our little Savannah hunting brains are just not up to that. So, you know, when you, uh, rearrange your, like when you move and your kitchen has the silverware drawer in the wrong place, like it's a better place in this kitchen, but it's still different and you go to the wrong silverware drawer for like a month.
That's, that's our failure to adapt, to rate of change. And so, um, you can't really like, gradually move the silverware across the kitchen, but we can absolutely
Coté 51:33
do I, that part, I that idea. That'd be funny.
Hedi Waterhouse 51:36
I think that would be funny for one person. I think that as a family, that might be possible architecture.
Coté 51:43
Now
Whitney Lee 51:43
there are for forks over here,
Coté 51:45
it's back to the needing group consensus to make change. You can, you can't just have three silverware drawers, you know, you're, you're making, you're, you're, you're making, uh, uh, this is reminding me of, uh, you know, the, you, you ever see that show Patriot? It's very shortlived show, but at one point this guy has to swim across the English channel through some electric ELs. And he, he makes this point of like, all you have to do is swim 51% of the way because then it's shorter to finish it than it is to turn back. And so it's almost like getting, getting to that 51% is, is like how you gradually get someone there that they put in the effort to do the rest. Is is another sort of, I don't know, hack ish thing to do.
Hedi Waterhouse 52:23
Right?
Whitney Lee 52:25
One thing I think is, oh, I was gonna
Hedi Waterhouse 52:27
change,
Whitney Lee 52:28
lemme finish your thought. Mm-hmm.
Hedi Waterhouse 52:29
I was just gonna say we need to make it internally rewarding. Like the thing about like, you only have to go 49% further is, um, you, we feel like that in software all the time. Mm-hmm. Like, okay, well I guess I have to do it this way now and I'm going to commit to it and I hate it. And in two weeks you're like, oh, what? No, I don't even notice it anymore.
Coté 52:51
Right, right.
Hedi Waterhouse 52:51
But that's because we have to touch it all the time. Mm-hmm. And I think that we are underestimating the fact that most people don't touch their bank account every day. And so if you make a change that feels month by month to you, it's still an abrupt change for a person who only logs in once or twice a month.
Coté 53:12
Hmm. Well, well, to to that point. You're let, let me ask you about this. I was just talking about this with someone earlier today. I, I often say I think all the time about this, but, but, but I don't think all the time about all the things I say I talk, think of all the time about, but often, let's say I think about like the pharmacy down the street and like their software needs or the, the goodness of them.
Like if they, if if the, if the, the pharmacy down the street were to like peg out in a good way, their Dora metrics mm-hmm. I'm not sure it would be too valuable to them because like I'm fine with getting my prescriptions. Like I don't really know if them like updating their software is like necessary, at least for the user facing software. Who knows what kind of. Happy hunting grounds of weekly changes they have for their internal facing stuff. But like at some point it's, there's a lot of software that I use, kinda like you were saying where like, I don't really need it to change. Like maybe deploying security patches, like all these kind of things is fine, but really, like, just like Whitney doesn't want the UI to change, like I, I would prefer you not ship a new u UI feature every week, right?
Mm-hmm. Like, and, and so it does seem like there's from, from a lot of DevOps, e Dora and other things, I don't always come across that like, well when is it okay to have a bad, a bad velocity? Like calibrating on like, you should only ship every three months. That sounds just like some counterintuitive ignite talk you might see off days. Ah,
Hedi Waterhouse 54:47
so the distinction is ship as fast as you can deploy, ask for user acceptance on a different scale. So the difference is that, um, rapid deployment, so I'm gonna change my terms here, but, um, rapid deployment is good. We know that from the Dora metrics and we know that it makes systems more stable. We know that from SRE, we know that from platform rapid deployment is good. Rapid release, which is the moment that it engages with the user is bad because rapid release is asking people to have a very high rate of change. There's nothing that says we can't deploy things that are not yet released.
Coté 55:40
Mm-hmm.
Hedi Waterhouse 55:40
Or not released widely. We're doing canary test, we're doing, uh, you know, capacity test. We're making sure that everything works, but we are not asking users to log into something that changes every day or every week. We're saying like, this is the cadence.
Whitney Lee 56:00
So they're getting, so you are doing your like, one to 2% canary deployment is having a small set of users see changes mm-hmm. Um, until they get accepted. And the, the idea that one user is seeing every single change is unlikely when you're testing the small percents all the time. Mm-hmm. Um, I have something that's really interesting to me about your book. Your book has four authors and y'all have wildly different backgrounds. So I'm interested in like, so, um, I could let you say, or I could say like, James governor, he's an analyst.
Mm-hmm. Um, that's understating it, but we're gonna do that. Uh, Kimberly Harrison is a, is a sociologist, right?
Hedi Waterhouse 56:41
Mm-hmm.
Whitney Lee 56:41
Yeah. Focused on technology adoption. There's you, um, you said your background's in developer advocacy, right? And, and, uh, technical writing. Technical writing,
Hedi Waterhouse 56:50
yeah.
Whitney Lee 56:51
Yeah. And then Adam Ziman, who's a, a startup advisor and a VC advisor. So how did each of those backgrounds play into the story of your book? And honestly, how did y'all find each other?
Hedi Waterhouse 57:04
So three of us were at LaunchDarkly together.
Whitney Lee 57:07
Ah, okay.
Hedi Waterhouse 57:08
So everybody but James was at LaunchDarkly from about employee 20 to about 2022. So like, uh, I think, I think AZ left a little earlier, but, so we were working together at this like breaking edge of what is it that we could actually do with feature flags as a service and how is that even a product
Whitney Lee 57:36
Uhhuh?
Hedi Waterhouse 57:36
And having to think about that really drove a lot of our understanding of what people want and how acceptance works. One of the really interesting things that, uh, Kim and I worked on was a survey. How people are using feature flags. And when we came on to LaunchDarkly, we thought feature flags were sort of a front end thing, like it changed the look and feel of how you were doing deployments and, and configuration.
What we found when we did this survey was that people were actually using feature flags in a, a load-bearing way. They were using it for, um, routing, like packet routing. I'm like, what? Why are you doing the, like, surely we have other tools for that, but it was the best tool for doing the thing that they wanted to do, which was kind of weird. Um, and people were using it for entitlements and, um, account management. It was so surprising to us that the majority of feature flags were what we called long-lived flags or permanent flags rather than ephemeral flags that you think of with AB testing and then you settle on a, an outcome. But no, people were using it
Coté 58:54
absolutely
Hedi Waterhouse 58:55
right to do. I, I think entitlements is probably the easiest way to think about it. And, and we're like, oh, we're not making what we thought we were making. And that kind of kicked off our thinking about how progressive delivery was, about understanding what you're actually making when it contacts the market. Ah,
Coté 59:16
why So what, why were they using those, uh, what'd you call them? Permanent flags? What, yeah, what accounted for them doing that?
Hedi Waterhouse 59:22
Oh, they're super useful. So, uh, you know how LinkedIn has two settings where either you're paying them or they're telling you to pay them?
Coté 59:28
Oh, oh, I, I, I mean, I mean, why did they choose to use your tool in a way you didn't intend it to be used to do that?
Hedi Waterhouse 59:34
Oh, because it was really good at what it did, which was delivering completely consistent, um, experiences across a global scale without you needing to understand how the edge network computing stuff worked. Like all of that was behind the curtain.
Coté 59:54
Yeah, no, this, this is interesting. I mean, especially it in, in light of all the stuff we talked about, it's, that's an interesting case of like, uh, to use an old phrase like the, the street finds its own uses. Like Yeah, it does. Here, here's a tool that's that's easy to use. Uh, well, it must have been easy to use and well designed and then also fit, as you were saying, like a global need for like all sorts of weird services and. Instead of using it for its intended purpose, people are like, oh yeah, we can just use this to, uh, route different people to different things.
Mm-hmm. And then turn on features or entitlements, as you say, depending on who they are. Which, yeah, that I, I guess, I guess that happens frequently or maybe not frequently, but that's, that's, it's almost like, it, like, like, like, uh, maybe confluence would be replaced by something that has nothing to do with it originally. And, uh, yeah, that'll work out fine.
Hedi Waterhouse 1:00:48
And it's so fascinating to me because coming to understand that the problem people were having was not understanding what a feature flag was, but how to get it globally distributed, consistently, it changed the way I taught. You know, I, I spent a lot of time teaching people what feature flags were, and when I started, I had a lot of conversations that conferences, do you know what a feature flag is?
And by the end. Yes, they knew what a feature flag was, but they had implementation questions.
Coté 1:01:24
Mm. Right.
Hedi Waterhouse 1:01:25
And I'm like, oh, that's really interesting. And I guess this is how adoption works and, and you see it happen. There's like waves. And the problem with being us is that we're way out on the front of early adoption. We're trying to explain like really radically new things to people who are still in the early adopter segment. And we're like, this'll be cool. Let me tell you about EBPF. And the majority of people who are using software daily are like. Um, I would like it if there were no backhoes between me and Chicago. That would be super. I don't, I don't know what happens other than that.
Whitney Lee 1:02:19
So I have one, one more question. Mm-hmm. Um, and, and I think it's gonna really pull everything together for me. Uh, I have high hopes for with no pressure here, but like, you're, you're basically presenting a new way to think about, um, software delivery lifecycle, and mm-hmm. A new segment of it. And so I, I would like for you, the book has some case studies. Will you tell like a, a concrete example of a company adopting this user feedback and how it affected them?
Hedi Waterhouse 1:02:56
So I. A really interesting example was the Disney case. And what happened is Disney has these magic bands that you can use to do all sorts of interesting things at the park. They're basically an identifier slash entitlement device, if you think about
Whitney Lee 1:03:15
it. Okay? So it's like a, something you wear in your wrist that
Hedi Waterhouse 1:03:17
gets access, you wear it on your, gives
Whitney Lee 1:03:18
access a thing, uhhuh,
Hedi Waterhouse 1:03:20
and it, uh, you know, keeps track of your credits and it helps you like hold your place in line. And there's a bunch of other things. When they built them, they built them with sensors that they did not have a use for yet. Okay? They did not have a software application in mind when they put in, um, an accelerometer.
Whitney Lee 1:03:45
Okay? What, what's an acceler accelerometer measure how fast you're,
Hedi Waterhouse 1:03:50
it can, it can tell how fast you're moving.
Whitney Lee 1:03:52
If you're on a rollercoaster versus
Hedi Waterhouse 1:03:53
walking, it can tell if you're on a rollercoaster and they're like, we don't know how we're going to use this yet.
Whitney Lee 1:03:58
Mm-hmm.
Hedi Waterhouse 1:03:58
But the hardware takes so long to build and develop that we are going to put as much as we can in and then develop software that we'll utilize this as we see needs.
Whitney Lee 1:04:11
Okay.
Hedi Waterhouse 1:04:12
And so the adoption for the magic band runs on a different pace than the software development for magic band.
Whitney Lee 1:04:21
Mm-hmm.
Hedi Waterhouse 1:04:22
So when you combine those, you can say, Hey, um, you've just gone on a bunch of roller coasters. Uh, did you know that there's a short line for this next roller coaster?
Whitney Lee 1:04:36
Ah,
Hedi Waterhouse 1:04:37
right.
Whitney Lee 1:04:38
Uhhuh
Hedi Waterhouse 1:04:39
or. You, uh, went on two rollercoasters and stopped. Would you like a, a meal? Like here's a little meal discount.
Whitney Lee 1:04:48
Yeah.
Hedi Waterhouse 1:04:49
Right. So
Whitney Lee 1:04:50
that's where the Dramamine is sold,
Hedi Waterhouse 1:04:52
right? And so building a device that you don't know all of the things you're going to use it for, it seems like a real leap of faith. But in a world with progressive delivery, you can say, software is not the hard part. I'm going to roll that out as mm-hmm. It becomes relevant to what people need or want.
Whitney Lee 1:05:12
Mm-hmm. And, and they're collecting user feedback as part of that,
Hedi Waterhouse 1:05:17
and they're collecting user feedback and they're like, okay, here are the busy times for this, uh, for. For and because it's, you know, an identity thing, they're like, okay, everybody with little kids gets here early and we need to have extra princesses out. But everybody with teenagers doesn't roll in until like 10 o'clock. Mm-hmm. So like, we're gonna have more capacity for the like scarier rides.
Whitney Lee 1:05:44
Right on.
Hedi Waterhouse 1:05:46
And
Whitney Lee 1:05:46
very cool.
Hedi Waterhouse 1:05:47
You couldn't really track that just by watching the ridership numbers. Right. You have to know who the writers are to be able to correlate that
Whitney Lee 1:05:58
uhhuh. That's awesome.
Hedi Waterhouse 1:06:01
Yeah.
Coté 1:06:03
Well, you know, you know what we do have correlation around here for is, uh, we've had a great discussion and hope, hopefully people have, uh ha have enjoyed it. If, if so we, we know, we know you've got the Progressive delivery book, but if people wanna check out more stuff about you, maybe look at that book, see what you're up to, where would you point them on the worldwide web.
Hedi Waterhouse 1:06:21
I have a website@heidiwaterhouse.com and we have a website@progressivedelivery.com.
Coté 1:06:30
And are, are you guys on a like year long book tour going on the conferences, talking about the stuff?
Hedi Waterhouse 1:06:36
Yes. I think our next appearance is, um, monkey Gras.
Coté 1:06:41
Oh yes. Oh,
Hedi Waterhouse 1:06:42
fun.
Coté 1:06:43
Yeah. Yeah. Cool. Are you gonna be doing any like readings? Does that, did you do that with this kind of book?
Hedi Waterhouse 1:06:49
Uh, not usually it's, but we do give away stickers so
Coté 1:06:52
even, well there's that e even better than a reading. You can read it yourself, but you can't make a sticker yourself. I mean, I guess you
Hedi Waterhouse 1:06:58
could. And there's an audio book. I am a big fan of audio books for technology books so that I can do something interesting. Well, I'm learning.
Coté 1:07:06
I just, just kind of ambiently pick it up.
Whitney Lee 1:07:10
Yeah. I'm laughing so hard at the implication that what you're learning is not interesting. I, I
Hedi Waterhouse 1:07:16
have difficulty reading technology books. I'm just like, Uhhuh Uhhuh. Uh, yeah.
Coté 1:07:21
I, I I could see how it would be kind of like, like yard fertilizer. It's like, yeah. Valuable but not interesting on its own. Yeah. I don't know. I'll have to work on that. Well, uh, speaking of, I know things that are valuable, not interested on its own, this episode of Software defined interviews, right, Whitney.
Whitney Lee 1:07:39
But, um,
Coté 1:07:41
I think, I think this is episode one 20, so if you wanna get links to, to, uh, things we mentioned and stuff, uh, you know that that site, Heidi said other relative things. You can go to software defined interviews.com/one 20 and find it all there. And, uh, with that, tha thanks for being on. It was fun.
Whitney Lee 1:07:56
Thank you so much, Heidi. Yeah, thank you.
Coté 1:07:58
Yeah, and we'll see everyone next time. Bye-Bye.
Whitney Lee 1:08:01
Bye.