(2023-11-14) Brian Chesky's New Playbook
Lenny Rachitsky: Brian Chesky's new playbook. Today my guest is Brian Chesky. Brian is the CEO and co-founder of Airbnb. I was very lucky to get to work with Brian for many years, and my sense is if you ask people who they consider the most inspiring tech or business leaders today, Brian would be right near the top of that list.
I want to start with the elephant in the room for a lot of listeners of this podcast. What is going on with product management at Airbnb? You made some comments at Figma Config and a lot of people got this impression that you eliminated product management at Airbnb. ((2023-09-09) Itamar Gilad On Linkedin Why Did Airbnb Kill Product Management)
I saw somebody tweet that said something to the nature of that I said I got rid of the product management function and all the designers in the room started cheering
It wasn't the people, it's the way they're working together.
we didn't get rid of people helping drive the product. What we did is we combined what one might call the inbound product development responsibilities of product manager with the outbound or marketing responsibilities of product marketing. (PMM)
The last thing we did is we made the group smaller and more senior. So we don't really have a lot of junior product marketers.
So the basic idea is this. You can't build a product unless you know how to talk about the product. You can't be an expert in making the product unless you're also an expert in the market of it
if you say you tried and it didn't work, my question is, was it a bad product, a bad strategy or bad execution? Maybe it was a really well-made product, but you had no distribution plan. You had no way to talk about the product. If you build a great product and no one knows about it, did you even build a product?
The other thing though is that they do not control or drive designing or engineering. We are a fairly purely functional model. They manage by influence.
Here's the amazing thing. We built and designed a company where you can manage by influence and no one has to like you. You don't actually have to win people over.
I think designers in the valley are very, very frustrated with the product development process.
I think a lot of designers feel like they're compromising.
design is catching things before it goes out the door. It's not actually typically part of the development process. And I think this is not just bad for design, it's bad for product managers and engineers because we all want to build the best products
The first thing you notice is that these different groups might be running on slightly different technical stacks. That's the first problem. And they may actually be accumulating technical debt. The next problem you'll see is that there's a lot of dependencies. So five teams are going different directions, but they all need a payment platform
instead of five teams going to marketing to get a campaign or to leverage some service, they start building their own marketing departments, their own groups. So now they're really becoming separate divisions. And this is where a division comes from.
now you have to advocate for your division. So there's a lot of advocacy. If you have dependencies, you've got to persuade people by building relationships
that creates what we call politics.
then you end up having this situation where a company's done 10 marketing efforts, but no customer's heard anything.
Marketing and engineering don't talk to each other. It's not even they hate each other, they're in different universes. I've always said that the health of an organization, one simple heuristic is how close is engineering and marketing?
And marketing is at lot of companies are like the waiters, engineers are like the chefs and the chefs yell at the waiters and they go in the kitchen. In fact, the waiters are the ones talking to the customers all day
my co-founder who obviously you know well, Joe, used to have this metaphor of lasers, flash bulbs and chandeliers. If you want to light up a room, performance marking is a laser. It can light up a corner of a room. You don't want to use a bunch of lasers to light up an entire room. You should use a chandelier and that's what brand marketing is. But if you do need to laser in and balance supply and demand, then performance marketing is really good
Performance marketing though doesn't create very good accumulating advantages because it's not an investment.
you actually need to be investing. And so we think of marketing as education. That we're educating people on the unique benefits.
So a lot of companies don't do product marketing. They do brand marketing, which are ads about the app or they do performance marketing, but they're never really educating people about new things they're making and shipping
And so you try these big new things, people don't adopt them immediately, so then you get more and more incremental
we have a rolling two year product plan. Strategy. Product strategy roadmap that gets updated every six months with releases. We release products every May and every November or October.
product management also does the product marketing. So they're figuring out how people are going to learn about it, they're doing the demos, they're understanding the story, the videos, they're figuring out all the customer touchpoints, making sure everyone understands it. Our product marketing works with communications. We work months ahead of time on all the different assets.
story also is a really helpful way to develop a cohesive product. We wanted a company where a thousand people could work, but it'll look like 10 people did it. (coherence)
So I'm going to pull on a couple threads. The first is this idea of a single roadmap you talked about
CEO of a public company and he pointed out that there's this cycle
you realize I need to take the reins again and drive the ship and take back control of what's happening
Brian Chesky (00:16:08): That's exactly how it went... many years ago, I remember I think reading a blog post by Ben Horowitz saying that a lot of people tell product led founders or engineering led founders to step away and delegate their product to other people, but suddenly they've delegated away the thing they're best at.
So we don't have a chief product officer (CPO) title, but if we had one, it would be me.
I think the CEO should be basically the chief product officer of a product or tech company. If the CEO is not the chief product officer then I don't know if they're a product or tech led company.
So when we were starting Airbnb, it was probably the three of us.
Let me be clear. There were times I inserted myself and dysfunction occurred. That is absolutely true. And that was just a learning experience for me. But there was this other scenario where the less involved I was in the project, the more spin there was, the less clear the goals
the slower they moved, the more they assumed it was because I was too involved.
We don't test blue versus green... if we do an AB test, there has to be a hypothesis
And then imagine 10 teams doing AB testing.
you have to think about the whole cohesive system
It was now late 2019. I don't know what to do. I'm like, the product is slow, the app seems to not change, costs are rising. I keep adding more people. There seems to be more politics.
And then right before the pandemic, I meet two people that really affected how I thought about things. The first was Hiroki Asai. Hiroki now works at Airbnb. He's one of my executives and actually product design, product marketing design and marketing report into him. And he was a creative director for Apple. So he worked for Steve Jobs, was basically dotted line to Steve for many, many years.
In Apple Marketing Communications, they actually designed the app. They made the app. They designed all the marketing touchpoints for the store. So it wasn't just ads, it was every brand touchpoint they were responsible
I met another person or got reacquainted a person named Jony Ive. And Jony Ive was the head of industrial design and Chief Design Officer at Apple.
And they described this way of running a company that was totally different than the way that I was running it. It was basically the way that Steve Jobs ran Apple from about 1998 until he died in 2011.
We had 10 divisions. We had a flights division and we had a homes division which was divided to pro hosts and core hosts and lux and we had business travel and we had a magazine and we had experiences and we had .org and we had China. So we had these 10 different divisions all going in 10 different directions.
And so the company kept getting subdivided, subdivided, subdivided and that made it very, very difficult to turn.
we lost our design roots.
And then the pandemic occurred and we lost 80% of our business in eight weeks.
And so suddenly I basically got into action and I said, I'm going to run it this other way where I'm going to get back involved in the details.
The first thing I did is I said, everything we're doing has to be written down and put into a Google sheet. It turns out people couldn't even write down everything they were doing. I remember one person told me, "We're doing too many things for me to ever be able to document." I'm like, "What?" (Org Writing Practices )
And I said, okay, we can do about 20% of these things.
instead of one team doing three things, three teams should do one thing
We removed layers of management
So we said, we're not going to have divisional leaders. We're going to have design, engineering, product which turned to product marketing and marketing and communications and sales and operations, all the functions of a startup. I said, we're going to have fewer employees.
We have fewer than 7,000 employees today. As a relative comparison, I think Uber has 30,000
we made sure that every executive was an expert in their functional domain
A design leader's job should be managing the design first, the people second.
How do you manage the people without managing their work? How do you give them development if you're not in the details with them on the work?
I stopped pushing decision-making down. I pulled it in. I created one shared consciousness and I said, the top 30, 40 people in the company are going to have one continuous conversation. Metrics are going to be subordinate to the calendar. So we're going to have a roadmap. It's going to be a two-year roadmap
I took a lot of product managers, I reassigned them as program managers. I had many of them trained in actual program management because their roles got much bigger. Program management at Airbnb is a high status job.
And then I had a head program manager that would score all the projects. Either they're green, yellow, or red. Meaning they're on track or not on track to ship. Whether we thought they were work, we don't know until after we ship it. (great let's bring back estimating)
But I use the reviews of the work every single week.
because we had all these reviews, I didn't need to mandate people going back to an office. I didn't really care where they worked because I could track how well they were working because of the review cycle
We built our own in-house creative agency. So we use production partners, but we don't use Wieden+Kennedy or ChiatDay or any of those anymore. We actually built our own in-house agency
We got rid of a function called UX writing, and we combined it with marketing writing. We said, wait, why don't we just have the best writers do everything? (Not the same kind of writing. Marketing writing is rarely about clarity.)
I think that way too many founders apologize for how they want to run the company. I don't know why they do, but I think they apologize for how they run the company. They basically find some midpoint between how they want to run a company and how the people they lead want to run the company
I basically got involved in every single detail. And I basically told leaders that leaders are in the details. And there's this negative term called micromanagement, and I think there's a difference between micromanagement, which is telling people exactly what to do and being in the details
Here are the things I believe. I'll give you a checklist
Number one, I think that the CEO, unless they're not a product person, should think of themselves as this chief product officer and they should be involved in the product.
we still spend money on performance marketing. We still do measure conversion and we will do some experiments. Think of conversion and growth optimization as like running a football down a field and think of these big leaps as passes. You should probably be doing 80% passes, 20% running the ball down the field, and a lot of companies that do 80% running down the ball down the field and 20% passes
Number two, if you're not functional, I would at least think about everyone being really close together.
I think that every leader should be an expert in what they're leading.
You have to know the subject matter.
Five teams should do one thing rather than one team do five things
I think that people should consider doing launches. You can by the way, ship every hour of every day, but then package it and tell a story if you want hold the product back
If you're going to do AB experiments or measure data, you have to understand what it means.
I would make sure that you have engineering and design ideally report to the founder, product led person. I would not have design under product unless you have an extremely good reason or the product person is a designer.
I would try to think about product management, expanding the responsibility and including distribution. Understanding the customer and teaching people how to tell a story
I would make sure that you have as few layers between a CEO and other people
I think you should think of each release as a chapter of a story or like an episode of a TV series
you had a huge launch today. I know you wanted to talk about it. Your winter release
The problem is that every home is one of a kind and you often don't know what you're going to get.
And so what we found is that reliability is Airbnb's achilles heel
so we asked ourselves, what if we could combine the uniqueness of Airbnb with the reliability that you've come to expect in a hotel? And that's what we've done with Guest Favorites
create the top two million homes (out of 7mill, that's 30%, seems like over-pitch)
another point I might bring up, is we've completely overhauled the host tab. One of the most important things when you get to an Airbnb is the listing is accurate. But the problem is that a lot of host listings don't have all the details up to date.
when you were at Airbnb, we had a guest team and a host team. We don't have a guest team and host team. We have a design team. We have a marketing team.
almost everything involves connecting the guest and host. When you have separate teams, they tend to have separate roadmaps that go in separate directions that become incompatible
By the way, the design is a whole new aesthetic. I'd like to make the announcement that I think flat design is over or ending.
We have this incredible new tab called the listing tab that we designed. It's quite possibly one of the nicest things we ever designed
I think the reason why is we're spending more and more time on screens and we want the screens to replicate some of what we see in the natural environment. Light, texture. I think it's more intuitive, it's more playful.
There's seven million listings.
we also felt like if we put care in the design of our app, that hosts are going to see that and that they're going to actually put care into hosting
Speaking of great products, a defining characteristic of Brian Chesky in my mind is how big you make people think. How you push people to think bigger.
Memories I have of you is in meetings, we present our goal and you're always saying, "How do we 10X this?
What have you learned about just the power of setting really ambitious goals, but also finding the balance with not demoralizing people if they don't hit these really ambitious goals?
As you know, there was a saying inside of Airbnb, it was add a zero
It's more the exercise of what would it take to be 10X bigger or do something 10 times better?
to think differently about the problem means you have to deeply understand the problem. And to deeply understand the problem, you have to break it into its components.
The second is, I think one of the most important things for a founder or leader to do is set the pace of the team. I think the pace of the team is one of the most important things you can do. And that pace is sometimes governed not by how hard people work, but how decisive they are
If we're in a meeting, we don't just say, "Okay, let's circle back on this next week." No, we'll have it done by next week. Let's stay in this meeting until it's done
What have you learned about avoiding burnout and creating balance
The more in the details I am, the more time I have on my hands. That's a paradox.
If you decide to be in the details and get very, very hands-on like I did, it might be a lot more work for about one to two years.
But once we turned the corner, suddenly everyone started rowing the same direction. Suddenly I didn't have to be in meetings anymore and people would do what I wanted them to do if I wasn't there.
that's what the culture is. They say the culture is what happens when you're not in the room
Before, I would get 10 surprises and nine were bad. Now I get 10 surprises and nine are good.
I do not think I am the poster child, at least historically, of work-life balance. I'm 42 years old. I live with a golden retriever. I don't yet have a family
sometimes you need ... Let's say an artist, you have to step away from the painting. And you actually start getting more derivatives, slower and slower.
I now try to say no to what I call fake work, which is things that feel like work, but they don't actually move the ball down the field.
When I was five years old I remember my parents take me to the Norman Rockwell museum and I would sit
I went to RPI when I was in freshman year of high school to do a pre-college program. Then I got into more and more drawing and figure drawing
And I was at this private school actually for hockey.
And I had an art teacher in high school at this military school. That's another fun fact. I basically went to a military high school.
And so I leave this high school because I wanted to pursue different interests than hockey. And I transferred to my public high school late my junior year. And imagine transferring to a new public high school late junior year.
I ended up being a winner, there were multiple winners, of a national art competition, and I had my artwork displayed in the Rotunda Gallery. That then led to me getting a scholarship at the Rhode Island School of Design where I ended up going to RISD.
I realized I was born 100 years too late for what I wanted to do, which is draw and paint. And I felt like at that point, photography and now AI generated art, but certainly even back then, photography was replacing a lot of the skills that I had
And that's when I was in my freshman year of college and I learned about a field called industrial design.
I don't think I could have ever done what I did if I wasn't an industrial designer. I think industrial designer is different than a graphic designer. Because an industrial designer, you have to work with engineering in your training.
You have to understand manufacturing. Industrial design's very accountable to sales
So you had to understand marketing and strategy.
But I didn't really want to make items and objects my whole life
there's not a lot of designers or artists running Fortune 500 or S&P 500 companies. And I think that intuition, imagination, design, curiosity, I think we need more of that
I think that some of the greatest scientists played musical instruments like Einstein. I think that being a whole, well-rounded way of thinking about the world is good.