In episode 13 of Unintended Consequences, Heidi Waterhouse and Kim Harrison continue their conversation with cloud economist Corey Quinn. This time they’re unpacking The Duckbill Group and how Corey manages to keep up with his Herculean content schedule.
About the Guests
Corey Quinn is the Chief Cloud Economist at The Duckbill Group. He hosts the podcasts AWS Morning Brief and Screaming in the Cloud, and curates Last Week in AWS, a weekly newsletter summarizing the latest in AWS news, blogs, and tools.
Corey Quinn: Yeah. Computers are super annoying.
Heidi Waterhouse: They are. They require constant reconfiguration every time something new comes out.
And I used to love new phone day. I'm like I got a new phone.
I get to change the wallpaper and do all the setting.
Now I'm like, "I hate everything." Somebody has moved my cheese. No.
Corey: Yeah. Today's new apple watch day, which is exciting as it gets.
And it really sort of screwed me over earlier, because I saw a couple people announce that they were starting jobs at AWS.
And then I congratulated leaders at other tech companies on their new hire.
And it turns out that my watch is two years fast.
Heidi: Oh, I thought it was only 18 months. But yeah.
Corey: It really depends. I've done exactly what the drift factor is on this, but it happens.
I'm not saying they have a retention problem, I'm just saying I don't make long term plans around people still being there.
Heidi: When I was living in Seattle and working as a tech writer, I had a strong theory that they just fed all of their writers to a Kraken.
They were always hiring 20 positions, no matter what happened.
And I was like, "What are you doing with these writers? How are you growing so fast that you never have a gap of less than 20 writers?"
And I said this at a tech writing conference, and the guy leaned over to me and says, "I am the Kraken, lots of them quit."
I'm like, "Well, I'm not working for you, but good to know"
Corey: Oh, we burn through people like you wouldn't believe.
I mean, at some point a scale, talk about Unintended Consequences.
You have the Amazon problem where they're running into population limits, as far as who hasn't worked here yet that they can burn through.
I mean, at this point only a fool would trust buying birth control from that company just because they need to increase the population to maintain their employment practices.
I'm just waiting for a horrifying announcement about changes in policies around fraternization among employees.
Kim Harrison: I think I saw this movie.
Corey: That gets dark quickly, doesn't it?
Kim: A little bit, a little bit.
Heidi: But it's true in places with a strong Amazon presence it's like, have you worked there yet and burned out or are you still shiny and optimistic?
And I'll see you in a couple years. And it's funny, working in Seattle.
You're also dealing with Microsoft and lots of people used to work for Microsoft, but never got to work for Microsoft because we were all orange badges.
And therefore a lesser cast of human.
Corey: That's part of the problem is I've been a consultant in a lot of different places and a lot of different companies.
And what I always found vaguely amusing was, "Oh, you're a contractor. So you have to pay for a coffee, but employees get it for free. Yeah. I'm also there as a contractor and we have an agreement, so everything gets reimbursed. So all right, fine, effectively it's a convoluted mechanism in which I still get free coffee and one American express membership reward point for the dollar coffee."
And that's great, but it's about unsellable being performative.
We don't necessarily have to treat our employees better in so far as we can get away with treating our contractors worse.
And okay, there are optics to this that I get it. I'm not here to try and change the industry, but I will comment on it.
And eventually some companies do seem to figure it out and start to realize, "Huh. Well, if we have contractors in, then I guess we might want to consider, I don't know treating them like people."
What a concept. I'm talking actual contractors.
We've decided to basically float employment law and reclassify people incorrectly.
Heidi: Yeah. Well, so that was the interesting unintended consequence that happened to me.
I worked on the BitLocker team very early on. And I did their first year of documentation.
And then my contract was up, and they didn't have the head count for a permanent person.
They couldn't hire me. But because they had the lawsuit about the people who had been permanent contractors for years and years, you couldn't work for them for more than a year.
You had to take three months off.
And so I just walked out with all of that knowledge and they had to figure out how to replace me.
Corey: I've never been a big fan of consulting or contracting engagements or models where, "Oh, I'm going to be embedded at my client for 30 or 40 hours a week."
It's yeah. That's not running your own business.
That is effectively being an employee without the protections afforded to you thereof.
I mean, it was important to me when I started this place that I not have any one client be an outsized proportion of my revenue.
And of course, being your first client, then all right, you're 100% of my revenue, welcome aboard.
But I've gotten to a point now where I think the largest one is still low single digit percentages.
It took work to get there. But that is the sign of a sub-level healthy business, as opposed to...
But yeah, 80% of our business is this one company.
And at some points why don't they just acquire you? It's a lot easier.
Heidi: Yeah. So tell us a little bit about how you manage Duckbill?
How do you know about the cool new podcasting tools and how do you manage the fact that if you're doing that you have over a dozen clients, all of whom may or may not have different expectations.
What are the mechanics of that?
Corey: Dozen clients at any given time. We had a lot it turns out.
But the answer is that I am the very loud, arguably obnoxious face of the company in many respects, but we're 10 people that are employees.
And probably about that as far as very specific contractors go, I'm talking actual contractors, not misclassification again.
And like I have HumblePod is a company that they produce all of our podcasts and they do a number of our video nonsense things as well.
Chris Hill runs that, and he's phenomenal at this.
And I throw things over the wall to those folks.
And I don't worry about it.
Every once in a while, he'll reach out with a, "You should buy this equipment to upgrade your podcast studio."
And I don't even bother to ask why.
Okay, great. And it's because I trust him, I trust his work.
Same story with tooling that we use for different things and things that I encounter in the wild.
In that sense, it's I am not an expert in this stuff, going from being a single independent consultant to having a company means it's a long period of hiring people who are better at the thing you are hiring them to do than you are.
And it's a constant process of being humbled. I was never planning to take on staff.
So instead I took on Mike Julian, my business partner. He and I go back a very long time.
And we are opposites in many respects. Like his love language is spreadsheets.
But our values are the same. And that's what's important, what makes this whole thing work.
And he in fact does a lot of the management of people.
Whereas, I mostly am outward facing making a nuisance out of myself.
Heidi: I love that it's very clear that you have this whole team and you couldn't be you and loud and obnoxious if you didn't have somebody who helped you with scheduling and who sent you podcast equipment.
And I think it's easy for people who are looking at the shiny celebrity side to be unclear on how much work goes on under that surface.
Corey: My content schedule is punishing. I have three episodes of Screaming in the Cloud that go out every week.
I have another three episodes of the AWS morning brief, different podcast that go out every week.
And I have three issues of a newsletter that I write, and a full blog post.
A lot of content reuse, which is helpful.
But at the same time, it still means that I need to be very judicious in what I do now.
I'm prolific. I can write very quickly and it comes out mostly intact, but I have editors who are fantastic at what they do. I have people that do anything that does not need to be me doing it. I attempt to find a way to outsource that to someone else. I don't need to be the person that copies and paste the thing into the WordPress CMS. I can have people do that. I don't need to be the one that picks the fun image for it.
I occasionally do when I have a fun idea, but great.
Twitter is an interesting one, just because I don't let other people post tweets to my account.
The most it's ever gone has been that there's a permission to prove system, people can submit them using buffer or something like that, but I have to approve them and I stopped doing it just because it didn't work.
That's what the last week in AWS Twitter account is for. It's not a company account, it's my own.
And that means that if people don't like what I have to say, they can basically shove it, I don't care.
I can take feedback from people that I think are well intentioned and operating in good faith.
Internet randos might tend to take with a little bit of suspicion and distrust at first.
Now someone says this hurt my feelings. I'm paying attention to that.
If someone decides to come in and just be a jackhole. Yeah.
I make constant use of the block button because I don't need the grief.
Heidi: Yeah. Twitter is a whole microcosm of the unintended consequences of scale, like top to bottom.
There are break points where you have and number of followers, and it stops being usable, and it's normal human format.
And they keep trying to fix it, but badly.
Corey: It's a hard problem. I envy them. At least they're trying now.
There was a long time where it felt like it was frozen in the amber, where they didn't understand what they had and they were terrified to make any changes because they didn't know how they might make the magic stop working.
It's how I feel about how flight works.
When I'm on a plane, I don't think about the physics of it because if I think about it wrong, it might stop working.
Heidi: It's true. It's a big metal bird.
Corey: Oh yeah. That abducts people.
Kim: So you are heavily involved.
I mean, you are leaning on a fantastic support staff, but you are heavily involved in driving these conversations and writing these pieces.
Where do you find time to also take on clients and provide feedback to these organizations that need help managing everything, AWS?
I mean, where does that come in and how do you make space for that?
Corey: Well, remember I do have a sales team. It's awesome.
I don't have to think about these things in the same way, I outsource that.
I very rarely do any sales work, which is awesome. I had to do all of it originally.
Again, you hire people who are better at stuff than you are. We have ahead of sales.
We have at the time of this recording, one account rep who is about to have a coworker in a week or so.
And that's just on the consulting side, media as a whole separate kettle of nonsense, it will lead out of this conversation for the time being.
I have a team of Clouds economists, which is how I pluralize it.
Tim Banks is our principle Cloud economist, and they are the ones that do the vast majority of the actual bill analysis.
I act as a point of last escalation on some level, I look at a really interesting in depth things that really require some insight and some visibility into.
Something I'm instituting now is once a quarter, I'm going to be doing one of the engagements entirely by myself, just so I could keep my finger on the pulse.
But it comes down to making the time for the things are useful and valuable.
Another type of our consulting engagements we don't talk about very much is a lot of hurry up and wait.
Specifically the ones where we negotiate on the client's behalf with AWS around contracts, large value deals, partnership agreements, et cetera.
It turns out that those things are fairly templated and things that we learned in one can absolutely apply to future engagements like that down the road.
And it's not what people think it is.
In fact, AWS is generally very happy when we're involved in this because everything just works more smoothly.
You don't have clients asking for things that are not reasonable or grantable in the same way.
And suddenly the data's a lot clearer and everyone's much more aligned, what a concept?
It seems counterintuitive, but there's no one on the other side of almost anything that I do, no one wants the AWS bill to be larger than it is.
And unnecessarily for getting value for it. No one wants companies wasting money on this stuff.
No one wants Cloud marketing to be boring and dry and ignored.
No one wants the entire space to be joyless. Have fun with it is my perspective on it.
Kim: Do you think there's people within AWS who secretly recommend, like you guys are asking for a lot here, have you heard of this guy?
He might help you out on other side.
Corey: Oh, I have a list of confirmed incidences of that happening.
Again, it's what we do aligns super well.
We are not AWS partners, but there's nothing saying that they can't recommend what we do and how we do it to their customers, when it makes sense.
As of this point in time, haven't done any outbound work as far as reaching out to people to say, "Hey, you want help with your AWS bill."
We've grown to the scale that we have over the timeframe that we have based only on inbound interest and referrals from existing clients, which happens a lot.
People are happy with what we do.
These days you don't have to go too far asking questions about, "Well, I have a bad AWS bill. What should I do?"
Have you talked to the Duckbill Group?
Kim: So what does it take to become a Cloud economist?
Corey: A sharp blow to the head is where it starts.
And seriousness as we envision it now, usually the background looks especially like SRE, because cost and architecture are the same thing.
And the finance aspect is something that is easier to teach to engineers and is to teach Cloud engineering to finance people, that's now.
Down the road I can see bifurcating that out as the team continues to grow, which we have plans to do.
And it starts to look very different.
But it's a combination of having the technical skillset, the analytical mindset and the ability to speak coherently to customers at a wide variety of seniority in place within the organization and levels of technical acumen.
In other words, you have to be able to present in a variety of small meetings to a broad array of constituencies.
It's a weird combination, but it works well.
Kim: It's interesting. It makes me think of developer advocates.
I mean, you're asking for two very disparate places that normally don't collide at a business.
Corey: In fact, I argue probably three, but yes, you're right.
It's hard to find. Historically, we look for people with consulting backgrounds just because they've at least mastered some of those things through the course of their career.
At least it's an indicator that they have.
And our interview process is extremely humane and also fairly effective at ferreting that out.
Heidi: Well, you heard it here, folks. First, the sharp blow to the head and then the consulting career.
Corey: And that sharp blow to the head often looks like a surprise AWS bill.
Heidi: Oh, I have a couple more questions that I want to get in before our time ends.
The first one is, what one thing would you do to Futureproof if you were running a brick and mortar hybrid right now?
If you're running a company that makes a thing, what would you be doing to think about the fact that software has eaten your world?
Corey: Remember, I'm a services company, myself. I'm not a software company.
And I've thought about this too. Because people say, "Oh, when are you going to expand?"
What happens if AWS rents a herd of unicorns and fixes their billing system.
When I started the company, I was worried about that.
Now I'm worried, they never will. It comes down to keeping an awareness of what the industry is doing.
And when what you're working on starts to deviate, you can dig into your heels out of the sense of this is my identity, no.
Or you can evolve of with it.
I was deep into running email systems at the start of my career, turned out that well, not every company needs a full on email admin and I saw that coming.
So all right, configuration management, shaft, puppet, salt, things like that. And then Docker arose.
Okay, great. Then I pivoted over to Cloud. Then I pivoted into these specifics of Cloud billing.
But if that stops being a significant problem or AWS is no longer what it once was.
And suddenly it is no longer the clear 800 pound gorilla in the room. It's going to be time to reevaluate that.
But because I'm not building software, that's taking me years to build here.
I can shift extraordinarily quickly compared to an entire industry.
And I expect this to be the last job that I ever have.
But that doesn't mean that I'm going to spend the next 20 to 30 years doing only what I do right now.
It's going to evolve with time.
No two days look alike for me here. And that's what I love about this.
If what I'm doing in five years, and give you the same answers I'm giving you now, something has gone awry somewhere. And that's not ideal.
Heidi: Yeah, that makes sense. So you'd say stay alert, be ready to pivot.
Don't double down on what was.
Corey: On some level, yeah. On the other, It's understand what your customers care about, because that's where it's going to come from.
I don't care what the analysts say. I don't care what Twitter says.
What I care about is the customers you serve.
Pay attention to what they have to say and the direction they're going in.
If you're selling data center stuff and they're saying, "Well, we're there to look at Cloud."
Arguing against Cloud is not the long term approach. It's self serving.
It's something that okay, they are talking about going to Cloud.
Everyone's talking about going to Cloud. Maybe that means it's time for me to look at what helping customers in the Cloud might look like. Or ultimately shifting in some other direction so that helping customers whom you have the relationships is suddenly irrelevant in where they happen to be hosted.
It really depends on the specifics of what you're building, how you're doing.
Now, if you're talking about a company that isn't selling software necessarily, and it's selling widgets instead.
Well, at that point, that's the widget market. They are experts on that. I'm not.
I'm the expert quote unquote "On a very narrow slice of things that touch AWS billing."
But I don't go into companies saying, "Oh, you're, I don't know, a social network or Twitter for pets or whatever it is or an antisocial network."
Great. And pretend I know more about their business than they do because I categorically do not.
That is, I think something that gets lost.
Where it's, I look at something that presents as a misconfiguration or a bill that should not be incurring some of the charges it is incurring.
And my question is, can you help me understand what's going on here and why this is the way that it is because there may be a constraint I don't see.
And showing up acting like I know what has shaped this better than they do is a move that just screams junior consultant who will not be here long, because it doesn't work out.
Heidi: Yeah. I can see that.
Corey: Honestly, if I want to be smarter than everyone else, I would just go work at Google. I can.
That is the corporate perception, but it is not aligned with the people I know who work there.
And it took me a long time to reconcile those two things.
They try not to give too much advice and recommendation for other people's contacts.
They talk about what they're doing instead. And that is often heard as we're Google, we know best.
It's a messaging problem. Who would've out that Google had a trouble with messaging, either in product or in marketing.
Heidi: It's a mystery to us all.
Heidi: All right. Who else should we be talking to?
What else should we go explore on this podcast to talk about Unintended Consequences?
Corey: Well, I don't know whoever set the pricing for managed Nat gateways at AWS would be a good start.
It's one of those things where it just becomes such an egregious price hog at some ends of the market, that it's one of those things that no one is happy about its billing model, but they keep effectively milking that cow until it dies.
So we'll see how it goes. Maybe somewhat of an abattoir, I don't know, learnt to kill the cow faster.
On some level, the unintended consequences are interesting at the macro scale too, because there have been a lot of parallels drawn between Jeff Bezos and Alexander the Great primarily in Jeff Bezos's own mine, but at some point there's that quote at Alexander wept for there were no new worlds to conquer.
And that's what happens when you hit a certain level, you have to start expanding into new markets all the time because the markets that we deal with, worship growth above all else, and okay, what's next?
I've never found that compelling.
I mean, people are going to do what people are going to do.
But on some level, the thing that I aspire to have is enough, never this insatiable hunger to grow at all costs.
And regardless of who or what I burn through and who and what I hurt, no thanks.
I'm not here to cast dispersions on people who make different choices on this, but I got to live with myself.
And that means different things to different people.
And for me at least, I want people to hate me for the right reasons, not the wrong ones.
Kim: That makes sense. All right. You read really fast. What are you reading this week?
Corey: Mostly a whole bunch of emails and outreach.
Because as we're recording this, we are winding down the last week in AWS charity t-shirt campaign, which has been doing well.
It's raising a lot of money for a good cause. 826 national, they teach, like I'm conflicted on this, because it's a youth program that teaches kids to creatively write, which is objectively a good thing to my mind.
But it also feels like on some level they might be doing them a disservice because they will be completely unsuited for a job in Cloud market once they enter industry, because creativity is not appreciated there from everything I can see.
Heidi: I guess we'll just have to hope they go into something more rewarding emotionally.
Corey: Absolutely. I hope so.
And when they wind up entering the industry and I want to be creative, but this thing sucks come and find me, talk to me.
I assure you, I will point you in the right direction. I'm kidding on that mostly.
But that's what I've been spending a lot of time on this week.
There's also a whole bunch of writing, mostly because I'm also taking my first vacation next week in, well, a couple years.
So I am going to unplug as much as humanly possible. I'm throwing things over the fence.
I will be reading a whole bunch of trash fiction next week hopefully, because honestly the AWS documentation is a bit of a dry read sometimes.
And I'm looking for like sci-fi and fantasy are great. My usual fantasy collection of things I read is press releases from Cloud companies.
But here in the real world, things don't always work that way.
Kim: This is why we need better writers. These press releases could be far more interesting.
Heidi: When I was a technical writer I used to say that I wrote speculative fiction, just for industry wages.
Corey: Yeah. That is a really good way of framing it. "What do you think of this documentation?"
"Well, I found the protagonist unrelatable is apparently not what they're asking for."
Heidi: Right. But I believe that in the future, it will be possible to install this using the sequence of stuff.
Corey: On some level, it's weird.
This is how you install the open source version of our product, note that it is 800 steps.
The enterprise offering is basically we reduce that and to click a button, please buy our product.
And then they make billions on IPO because, A, there's value in saving people time and effort.
Heidi: All right. Anything else you'd like to leave us with?
Corey: Just the usual. That if my sense of humor appeals to you, feel free to sign up for my newsletter, which goes about like this conversation has, @lastweekinaws.com.
Heidi: All right. Thank you, Corey.
Kim: Thank you.
Corey: Thank you, Heidi, Kim, it's always a pleasure to talk to you, while again, indulging the love affair I maintain with the sound of my own voice.