(2024-10-28) Airbnb CEO Brian Chesky Says

Airbnb CEO Brian Chesky says ‘founder mode’ came from studying Steve Jobs. One of the reasons I love talking to Brian is because he spends so much time specifically obsessing over company structure and decision-making — if you listened to his previous Decoder episodes, you already had a preview of founder mode because Brian radically restructured the company after the covid-19 pandemic to get away from its previous divisional structure and transition into a more functional structure that works from a single roadmap. ((2023-11-14) Brian Chesky's New Playbook)

let’s start with the news: the Winter Release. The last time you were on, you actually talked about staggering Airbnb’s releases on one timeline for the whole company, so you get the summer release and then the winter release

introducing something that we call the Co-Host Network... we have more than 8 million homes today, but we’d love millions more... we asked, “Well, why aren’t you hosting?” The number one answer people gave us was that they perceived it as being too much work.

There’s many companies that can manage your Airbnb. The challenge we’ve noticed is the average five-star rating for third-party property managers on Airbnb is about 4.62. It is significantly below the median range review score.

So we thought, “What if we basically put together a whole collection of Airbnb-certified hosts that can manage your Airbnb for you?” People who only manage a few properties, so they’re not going to be managing you and 100 other homes. Maybe three, four, or five other homes. The average rating of a co-host is a 4.86, 73 percent are super hosts, and we have 10,000 of them today

We also have more than 50 upgrades for guests. The basic idea is to make Airbnb a more personalized experience

you’re a third-party company, you don’t want to have a business with just managing five people

you manage thousands of properties, now you have a whole bunch of employees.

This is basically just a little more of an industrial hospitality experience.

is a little bit different than the original ethos of Airbnb of “living like a local.”

The part that makes Airbnb hard is that there are 4 million people per night from nearly every country in the world — more countries than Coca-Cola operates in — living together

it seems like professionalizing or certifying that class of user on Airbnb goes a long way toward ending some of the gamification (game-playing) that occurs in the platform.

There was an entire industry of companies that emerged — that we caught, and we stopped them — that were advertising that they can get your bad reviews taken down. They knew how to call customer service, and what to say to get your negative reviews taken down. That became not only a fraudulent activity, it became an industry. There were entire companies built around doing this.

the thing that makes Airbnb so difficult in this way — and probably more difficult than, say, a video platform — is its longer tail (long tail). On YouTube, people can game things, but generally, a disproportionate number of views go to certain videos. For Airbnb, that’s not possible

We thought about how the internet is like an immune system, and what you should do is give communities tools to moderate themselves: so flags, reports of suspicious behavior, but, most importantly, a review system.

It turns out that is necessary, but not sufficient. So over time, we’ve been in the business of managing more and more of the quality ourselves.

The part where you’re taking more control of the experience, because you think that’s perfectly aligned with the customer, at some point do you just end up with a front desk? Are you just running the hotel? It seems like that’s the farthest end of that journey.

I think we want to have almost all the benefits of a hotel while retaining the benefits of Airbnb. The benefit of an Airbnb is that every room and every home is different. There’s no SKUs. Hotels are such commodities. 90 percent of our hosts are still regular people.

But what we do want to do is match the hotel service as much as we can. There isn’t a front desk, but can we create a remote front desk? Can we have more 24/7 support? Can we use AI to level the playing field of the front desk, where AI can be immediate, it can be multilingual, it can adjudicate?

What you can do is you can train it on the corpus of like 100 policies, and then it can look at the last 100,000 times somebody complained about this, and figure out what the most likely resolution was that led to satisfaction for both parties,

What we’re probably going to have in the future are more standardized quality tiers. So there’s going to be a standard to list on Airbnb, and then there’s probably going to be — depending upon the vertical — different levels of certification or quality

I said earlier that “I think the founder mode conversation and Decoder questions have pure overlap.” The last time you were on the show, you described re-orging Airbnb. You restructured the company, you went onto this roadmap where you shipped twice a year in big deliveries, and you said, “I got rid of all of these middle managers, and I’m the product manager, and everyone rolls up to me.” And I thought, “Oh, this is...Chief product officer.

a year passes, and you give a talk, and I thought, “I’ve seen that talk before

Then, Paul Graham goes and writes a blog post called "Founder Mode".

it’s not like I’m suddenly running the company differently than a year ago. It’s just that now it has a label, and I think it makes sense to everyone

from 2009 to 2019, I ran Airbnb the way most tech companies run their companies. I didn’t know how to run it, so I hired people from Google, Amazon, Microsoft, and other companies, and they brought their processes with them

I remember that we were kind of a matrix management organization with 1,000 employees, and like almost all matrix organizations, it was hard to get work done.

I’ll give you an example. There was a creative marketing department that would have to create graphics for different teams... it would be like a three-month waiting list because they were inundated

So then the team said, “Well, give us our own resources. Give me a dedicated creative person,” and this could also be true for technology, finance, a legal team, any function

And this is when you start to divisionalize the company. Quite literally you’re subdividing it. This is where the general management structure comes from. It makes a lot of sense why this happens.

The problem with that, though, is that once you subdivide the company, the company starts rowing in different directions.

The next thing is that these general managers are incentivized, typically, on output goals that you don’t usually run on P&Ls. They run on impact or growth (proxy metric), so they have to advocate for as many resources as possible. This is what we call politics.

Because the company’s going in many different directions, oversight becomes more difficult.

If I could summarize founder mode in a couple sentences, it’s about being in the details. It’s that great leadership is presence, not absence. It’s about a leader being in the details

And one day you’re going to wake up, and you have 50-year-olds managing 40-year-olds, managing 30-year-olds, managing people two years out of college doing all the work with no oversight, and you have these four unnecessary layers. You have no experts in the company.

So, the antidote to this is to try to be as functional as possible. We are a functional structure. Functional just means expertise-based, not general management-based.

I generally think the CEO should be the chief product officer of the company

Your leaders shouldn’t just be “managers” (and I put managers in quotes), they should also be in the details.

Now, here’s the problem with the narrative being in the details. There’s a term for it, and it’s a pejorative. It’s called micromanager.

a lot of founder mode came from me studying Steve Jobs. For 10 years, I was at wit’s end. It wasn’t my first instinct to copy Steve Jobs’ leadership style, it was kind of a last resort.

A few weeks ago, I had dinner with his son, Reed Jobs, and I remember asking him what Steve’s opinion of micromanagement was. He had such an interesting perspective. He said, “Steve Jobs was in the details.”

He said he never felt like he was micromanaging because he was partnering with people on the details. I asked people like Jony Ive and [Hiroki Asai], who worked for Steve, and I said, “Do you feel like Steve Jobs micromanaged you?” And they said, “No, he didn’t.”

I think the distinction is that I remember one time an executive on my team asked, “Is this your decision or is this my decision?” And I remember saying, “It should never be either.”

By the way, you don’t need to be a founder to do that. You can apply founder mode to government, you can apply founder mode to a nonprofit

it’s not about being so-called autocratic because you’re not telling the experts what to do.

One of the most important things I do is have an executive team. You could think of them as C-level or SVPs. There’s about seven of them. Then, the next level are VPs, and there’s maybe 30 of them or 40 of them

Something I do that’s different from almost every other CEO in Silicon Valley (but I think [Nvidia CEO] Jensen Huang basically does this, Steve Jobs did this, Walt Disney did this, and Elon Musk did this) is I treat all the VPs as direct reports. Jensen got rid of the executive team. He just has 40 direct reports. That’s a little unwieldy for me. I’ve gone through that thought experiment of just having the VPs, and I can’t track everything.

But all VPs do a report. They report to me and to their executive, and I am the co-hiring manager. So instead of me hiring an executive, like a CFO who hires their people, I am the co-hiring manager. I do the kickoff, I’m the second interview, and, ultimately, I decide the final compensation for all the top people in the company, not the managers.

you skip-level and manage as many people as possible in a dual-reporting relationship.

I don’t think all companies should be functional, but I think they should be as functional as they can get away with

There is one very specific downside: The downside of the functional structure is you cannot do disparate things. But here’s something I wanted to dispel. It’s not true you can’t do as many things. It’s not true that it would be a slower-run organization. We’ve actually increased product development (velocity) by being a functional organization, but we’re like one flywheel. We can’t create three other flywheels that are disparate. (strategic context, alignment)

The only reason I believe a company should divisionalize is so that they could do truly disparate things that have disparate functional expertise.

there is no function called “business.” Maybe we can also call it revenue — business, revenue, commercial. There’s different names: chief commercial officer, chief business officer, chief revenue officer. They all kind of mean the same thing. The reason most companies don’t have this is because they have general managers, and the general managers play that role.

We found that the problem with that is, none of them are really experts in “business.” When I mean business, I mean revenue, the business model, the market size, what are the kind of dynamics of this market?

So, we decided to create a chief business organization with three functions. One is supply

Then, we have a business function, which is quite literally the business counterpart to product marketing

So they have a counterpart called the “business function.” With Co-Host Network, for example, we had a business person looking at the business model for Co-Host, we had a product marketing person who understands what’s the product, how are we going to market this, and why do people actually want this. Those two people have to go together. One reports to Hiroki — it’s product marketing, it’s more creative. One reports to Dave Stephenson, the former CFO now chief business officer, and they’re really the two parts of the continuum.

Then, you have a supply person... They work with international, the third group in the chief business officer org, to then take that playbook and bring it to all the different countries. And, of course, the product marketing has their own three-legged stool with design and engineering.

But it’s different from other companies where product really directs design. I don’t like that, and I caused a bit of a storm at a Figma conference. Basically, I got taken out of context and people thought I said I fired all the product managers. What actually happened was that I got rid of the classic product management function, I reassigned the most senior product managers to be product marketers, and I reclassified most of the other product managers as program managers.

By the way, most product managers in Silicon Valley aren’t actually product managers. They’re glorified program managers, but they’re not even experts at program management. That’s all they’re doing; They’re making sure the thing ships. That’s program management, that’s not product management.

The second downside to a functional organization is that it takes longer to start. Because if you want to just get something going, you need to get everyone organized. But everyone organized has a multi-year roadmap, so they now need to make room on the roadmap.

What I basically tell people is it’s harder to get something on the roadmap, but if it gets on the roadmap, we put the weight of the company behind it.

now we’re prioritized, now we only do things that are differentiated. And the governor is that we only do as many things as I can focus on

This doesn’t work if you come from management consulting, and now you’re a general manager who doesn’t really know the domain

I ask a lot of CEOs, “How do you make decisions?” I would say one of the most common answers we get is, “It would be best if I wasn’t making so many decisions, if my team was empowered, and all I was doing was breaking the hardest ties or making the biggest, riskiest investment decisions so they weren’t feeling that pressure.”

I totally disagree with that. I think I’ve heard Jeff Bezos say that more recently. He said, “My job is to make as few high-quality decisions as possible,” and I could not disagree more with that

The truth is there’s multiple ways to do something. I’m strongly advocating for this way. Funny enough, I’ve talked to many people that were early members at Amazon, and that’s not how Jeff Bezos ran it early on. (It's in a lot more businesses now.)

here’s the key thing: I believe you need to hire smart people. The paradox is that I believe most smart people want you involved, they want your partnership

there’s a scenario where I have more power, and therefore you have more power. It’s not like zero-sum, and I’ve wrestled control. I’ve found that the most talented people like my involvement as long as the involvement is constructive, it isn’t me telling them what to do and pushing them. It’s like, “What about this? What about that?” Steve Jobs went to Jony Ive’s design studio every day. He wasn’t telling Jony Ive what to do. He was discussing things, and they were debating and brainstorming, and it was a partnership. I think this is a really, really important framework.

Somebody might ask, “Wait, you’re in all the details. How can you be in all the details forever? That seems like it’s not going to scale, and someday you’re going to get tired.” And the answer is they’re right. Here’s what most people do: They hire smart people, and they give them operating freedom. They have no idea if they’re good or not because they’re not engaged enough.

I did a couple golf lessons once, and I had a golf instructor who literally coached me on every single swing. Thank God I had the golf instructor because if I just went on the golf course on my own and I swung 1,000 times, I would have had a really screwed up swing by the time the golf instructor got involved, and it would take even more work to retrain my swing.

My philosophy is you start in the details. You’re involved in every single thing. You hire great people, and you’re in all their details. Over time, once they develop muscle memory and they prove that they understand the system, then you can gradually let go.

Two years ago, I wanted to write perfect press releases not because I think anyone reads them, but because if you can’t put your ideas down in a clear press release, then you don’t have clear thinking (PR/FAQ). You don’t know why you’re doing something. Two or three years ago when we did press releases, we’d do as many as 70 revisions.

The most recent press release for the Co-Host Network I probably reviewed three times.

But it was the repetition in the details, that’s how people learned. They were saying, “That’s apprenticeship, you learn.” Even if you hire experts, they’re not experts at your company, they’re not experts at collaborating, they’re not experts in your business.

When you were on this show the last time, I asked, “How do you make decisions?” and you said, “Let me tell you a long story.” It was about various Airbnb controversies where you had to make big decisions. And you said, “I’m going to have to make so many decisions that I went to all my team and I said, ‘Here are our principles. Trust me to use these principles, and this is how I’m going to make decisions.’” Is that still the framework?

What I basically said was, “You want to make principled decisions, not business decisions.” Principled decisions are: if I don’t understand the outcome or if I can’t predict it, how do I want to be remembered? What do I think is the right thing to do? (value)

what I’m doing is, I am basically making a decision.

probably day-to-day, the most important thing they do is make decisions. It’s really important that people understand the criteria off which you’re basing decisions. …

I went to my board and I said, “I’m going to have to make 1,000 decisions. I can’t run every decision by you, so let’s agree on the principles and the framework for which I make all the decisions. Then I’m going to make all the decisions, and if something stands out, I will elevate it to you. I will retroactively show you all the decisions I made. If there’s something that’s huge about the company and one-way door, I will tell you ahead of time.

the way I run Airbnb is I review all the work — every week, every two weeks, every four weeks, every eight weeks, every 12 weeks. You might ask, “How do you have people do a report? How do they keep two people happy?” And the answer is, we’re all in the same meeting together. I usually have my direct talk first because if I talk, you’ll just agree with me, and then I’ll make the final decision. Most of my final decisions are just agreeing. Ninety percent of time I’m agreeing with what the team says, and 10 percent of the time I’m disagreeing.

Does the company know the principles that you use to make decisions? Is that published? Is that a thing you talk about?

There’s not a single list. There are core values. There are strategic differentiators

There are those general things. Actually, here’s another way of saying it. I don’t push decision-making down, I pull decision-making in. I think of the company as one shared consciousness. (Hive Mind?)

I don’t manage all of Airbnb — I manage the top 60 or 70 people. But I’m very engaged. I pull them in, and I create one shared consciousness. If you were to interview every single one of my senior people, I don’t think they would have pithy word-for-word verbatim principles, but they would be able to describe exactly how I think because we’re in so many meetings together.

there are many people that spend 10–20 hours a week in a room with me. It’s the same few dozen people going through different meetings, creating one shared consciousness, and being able to finish each other’s sentences.

You’ve talked about Apple a lot. They don’t have Steve Jobs anymore, they don’t have Jony Ive anymore. They have started a bunch of divisions to do all kinds of other things. Do you think that’s sustainable for them?

No

I think Tim Cook has done an extraordinary job, especially given the cards he was dealt, right? What alternative was there? Steve died.

Tim Cook basically took Apple from $300 million to, whatever, $3 trillion, added more than 90 percent of the market cap, and it’s done very, very well. He’s a great operator.

doesn’t mean that Tim did most of the work. He inherited the most valuable product of all time,

I don’t think that any other product has generated more profit — maybe not more revenue, but more profit — than the iPhone.

Tim was in an incredible period where what they really needed to do was maybe not invent a new product, but take the most successful product of all time and scale it, manufacture it, and make it completely ubiquitous, and he did that

We’re in the change industry, and it is very, very dangerous to not be constantly changing. If you’re a company that makes devices, the most important thing you need to do is make new tools and make new devices.

Succession planning is hard because the people that are great product visionaries are typically young.

Founders are allowed to manage people older than them because they’re the founders. If you’re not the founder, people just don’t want to be managed by somebody younger than them who’s maybe a virtuoso, a wunderkind, but they’re a little immature

The companies don’t want to take that risk, so they bias towards senior, grown-up, functional experts. But typically that function is not the product, and I think that’s a problem.