Host Intro (Gene Kim)

To set some context for the next talk, I want to make some observations about certain choices we made about the conference programming over the last 15 events. I had mentioned on day one that in 2014, we really designed the conference to be one about large, complex organizations. We wanted the conference to be not about the unicorns, the Facebooks, Amazon, Apple, Netflix, Google, Microsoft, but instead organizations that have been around for decades or even centuries. A little trivia fact: the oldest organization that ever presented was Her Majesty's Revenue collection service in the UK, founded in the year 1200.

But over the years, we've increasingly had speakers from Amazon Prime, Google SRE, Microsoft, Netflix, because there is so much that they can teach us. As a percentage of talks, the number of these talks from those companies is growing, and I think it's because it's becoming so evident that they're facing so many of the same challenges that this community does: challenges that require great leadership, great architectures to overcome issues around coordination costs, architectures that have outgrown the problems, culture issues, and leadership.

For all of those reasons, I'm ridiculously excited for this next talk. It was John Smart and the program committee who introduced us to Anna Steel, who leads the Ways of Working Center of Enablement for Community Support at Airbnb. I learned just how important the Community Support function is because they are the people responsible for helping customers when they have problems, whether it's a guest showing up at a property not being able to check in and get a key, or if the listing is not as described, or if they feel like they're in danger. And because Airbnb is a two-sided marketplace, Community Support also helps the listers. She is part of an organization of over 10,000 support ambassadors in 26 countries, and she will tell the amazing story of a problem she saw that she felt was incredibly important and what she did about it. She will be co-presenting with Marcus Braszell, now a partner at Sooner Safer Happier, coming from a long career solving these types of problems at banks and retailers. Here's Anna and Marcus.

Marcus Braszell and Anna Steel

Marcus Braszell: Viva Las Vegas. My name is Marcus, and I'm from Sooner Safer Happier.

Anna Steel: Hello, Las Vegas. I'm Anna, and I'm from Airbnb. Today we're here to tell you about Community Support and our ways of working journey at Airbnb. We're community-minded at Airbnb, so we've been so looking forward to sharing our story with the DOES community this week. It's also been amazing hearing and learning from all of you.

This is chapter one of our story, which is all about ways of working in a customer service operational environment. To get us started, Marcus, can you tell us a little bit about Airbnb?

Marcus Braszell: Thanks, Anna. We began working with Airbnb in 2021, and we all know who Airbnb is, but just so we're on the same page, here's a little bit of information. Airbnb was founded in 2008 by three founders, and by 2011 it was a unicorn, which is a privately owned startup with a valuation of over $1 billion. Today you can access Airbnb in 100,000 cities. There are 4 million hosts, there are 6 million listings on the platform, and there have been over 1 billion stays on Airbnb.

These are impressive numbers, but for me the most impressive thing about Airbnb is this host-like culture. You really feel welcomed when you're working at Airbnb, and it's a privilege to work there. Anna, you mentioned this is chapter one of our story. Where does it begin?

Anna Steel: Our story begins with Community Support. We're the team that provides customer service to our hosts and guests. We like to say that every day, millions of our hosts and guests are sharing meaningful experiences on Airbnb, and when you need our help, we're there to answer the call or the email or the chat with over 10,000 support ambassadors worldwide. There are over 100 million trips on Airbnb every year, but 87% of the time, people never need to contact us. We're there if they do, in the moments that matter most, ensuring that your experience is positive, memorable, and meaningful.

To put it in context, Community Support is an integral part of platform operations at Airbnb. The service we provide is our core product, but we work with many other teams, from hosting to technology to marketing, to develop the core services and policies and products that improve the customer experience.

As we've heard, Airbnb is an undisputed success story. It's been such a privilege to be a part of that journey, and in Community Support, we're passionate about our mission. So why did we want to improve our ways of working?

Improving ways of working began here. Over the years, Airbnb has scaled rapidly, and as we scaled, so did our complexity. Whilst we still very much have startup in our DNA, all too often it felt like we were running fast but failing to deliver on the outcomes we really wanted to see for our community, and move the needle on time to market or time to value. Over those same years, we tried to improve our ways of working, but all too often those improvement efforts just didn't stick. They were too top-down, too process-focused. They only resulted in more bureaucracy, further compounding the complexity we were already grappling with.

In early 2021, as a leadership team, we were really thrashing with this. We know that we want to be more agile. We don't want a rigid methodology. We're starting to play with this concept of what more nimble ways of working would look like for us, in part because we didn't want that capital-A, capital-T Agile Transformation inflicted on our teams. It's around this time that our then chief of staff happens to pick up a copy of Sooner Safer Happier, which really crystallized our emergent thinking.

If we were to be successful in our ways of working journey, we would need to focus on outcomes over outputs. With this focus, we started to play around with what the future might look like for us. We developed these from-and-to statements to better articulate and communicate the why of our ways of working effort to our teams. It drilled down into this central core: we wanted to feel like one team striving towards clear outcomes, guided by shared understanding, and teams feeling psychologically safe delivering incremental value.

So with that vision in mind, Marcus, how do we get started?

Marcus Braszell: Thanks, Anna. As Anna mentioned, we didn't want a rigid, one-size-fits-all approach, and here's how we got started.

We began with a discovery process to learn more about the ways-of-working landscape. This involved interviews with the business and a ways-of-working survey. We learned that Airbnb has an amazing culture. However, the discovery process also illuminated a number of challenges. We learned that the eNPS for ways of working was minus 71, which meant that people weren't very happy with how outcomes were being achieved. We also learned there was a work-in-progress challenge, or a WIP challenge. For example, there were over 30 priority number ones. The question, "Is WIP manageable?" in our survey had a response of minus 74%.

These headwinds and these verbatims really summed things up for me. The first two were headwinds, where people thought that operating in silos was a culture of not wanting to rock the boat. However, this is contrasted by these tailwinds, where there is an amazing collaborative culture at Airbnb and a really strong, deep passion for its purpose, which is very hard to replicate.

We also wanted to create alignment for the direction of improving ways of working. We set about together to create this vision. A couple of key callouts here: first of all, we want long-lived multidisciplinary teams working towards clear outcomes, and importantly, a psychologically safe environment where people can do their best work. We've since found this to be a really great rallying call, and it's still the test of time.

We also didn't want a big-bang, learn-slow, capital-A, capital-T, annual transformation approach, as we're looking to think big, start small, and learn fast.

One of the first steps we did was to set up a cross-functional multidisciplinary team, which in Airbnb we call the seedling team, with the following criteria. Firstly, it had to be strategically important. We knew that if it was high profile, it was more likely to get the air, sunlight, and water it needed for growth. Secondly, fertile ground: an environment fit for experimentation, fast learning loops, and ideally some quick wins along the way. Next, we were looking for a caretaker leader who would protect the seedling and help to remove impediments. This is represented by the glass house that you can see sitting around our first seedling.

Next, we wanted this to be a long-lived experiment, to prove this out in context, so we didn't want this to be a short-term project. We also didn't want to mandate an approach to improving ways of working, as we're looking for invite over inflict. And we began with early adopters.

We set about finding the curious early adopters to join our first seedling team. Early adopters are the people in your organization who say "hell yes" when it comes to new ways of working. These trailblazers have helped us to create the social proof to cross the chasm from our early adopters to an early majority, and here I'm referencing Everett's diffusion of innovation.

So there's a bit about how we got started. Anna, what do we do next?

Anna Steel: We created a fledgling Center of Enablement. This was as much to support those early seedling teams as it was to help engender that broader shift in culture and mindset that we were seeking. We did a few things. We started with regular communications with news and updates vertically and horizontally across the organization. We also launched a podcast, created a ways-of-working community of practice, and heavily invested in coaching and training on everything from business agility to Kanban to leadership development. Again, not focusing on any particular methodology, but rather on the patterns that would help our teams apply agility in their own unique context.

This all laid a groundwork that allowed multiple seedling teams to start coming out of that glass house, survive on their own, supported with the tools and new language that we were giving them. We saw some really exciting early wins with those seedling teams. Flow efficiency doubled, lead time halved, engagement within teams had never been higher. We were regularly seeing eNPS of 100%, and that came through in testimonials we were hearing from our teams. People were saying, "I've never felt so productive," and, "I don't want to go back to our old ways of working."

Our seedlings also helped put a spotlight on our system of work. For the first time, we started measuring things like work in progress, throughput, lead time. We hadn't measured these before, and it had been a big blind spot for us. As we started measuring them, we ran experiments, and that allowed us to reveal those systemic challenges, those impediments, the broader organizational challenges that we need to overcome in our context. But whilst challenging, we know that impediments are the path. At this point in our journey, we're confident we're on the right path.

This is also where the change got really hard for us. This is what we were seeing: our organizational structure and role-based silos made it really difficult to form those long-lived multidisciplinary teams, even in a protected seedling environment. Our traditional approach to portfolio management, with work passing through stages, was only reinforcing big upfront design. A combination of an output-centric culture with high WIP made it really difficult for us to see outcomes, or experiment, or work iteratively.

Against the backdrop of these impediments was just that challenge of changing mindset, of changing culture. There's a really great quote our VP of CS often references, which resonates with me when I think about that broader arc of change in a ways-of-working journey, and it goes like this: all truth passes through three stages. First, it is ridiculed. Second, it is violently opposed. Third, it's accepted as being self-evident.

To get to self-evident, you need the social proof, and impediments are a really important part of that story. With all things being clear in hindsight, here's what came into view for us. In an output-centric environment, transparency of outcomes was going to be a culture change for us. In a culture of good news, for impediments transparency, we needed to make it psychologically safe for our teams to share what was getting in their way of continuous value delivery. To form those teams, we needed to break organizational silos and dependencies rather than manage them.

It's at this point we're also seeing we need even greater leadership buy-in. We had tremendous support from the top, with our VP of CS as well as our director sponsor. But what we were learning is we needed all leaders to really lean in, role model, and most importantly incentivize improving ways of working.

So armed with these learnings, Marcus, where do we go next?

Marcus Braszell: Thanks, Anna. Armed with the insights from our seedling teams and central enabling, we set off for a leadership team off-site. This is a really pivotal moment for us.

It was June. The entire CS leadership team was gathered for a QBR, and we presented on the seedling team results. We also pushed for a workshop on value streams. We'd had amazing support from our sponsors up until this point, but we saw at this stage the entire leadership team begin to lean in and start to rally behind the change. It was an incredibly exciting moment for the Center of Enablement.

Here's what we presented. We stepped through the anatomy of a value stream and conducted a hands-on exercise. We found this really helped to create shared understanding for this concept. Here are some of the key elements. Firstly, we wanted business and IT embedded in the value stream. As Anna mentioned, we wanted to break dependencies, not manage them.

Next, we wanted clarity on who the value consumer would be for the value stream, be it an internal consumer or external consumer. We also wanted nested alignment from teams through to the strategy in terms of outcomes, with progress being transparently measured through the use of OKRs. Next was multidisciplinary teams whose primary identity would be to the value and the value stream rather than a functional silo.

We recommended a number of these patterns and practices, such as customer jobs to be done, a single backlog for the value stream, and all the capabilities needed to get it done embedded within the value stream. Next, we looked for these three leadership roles: the value outcome lead supporting the what and the why, the team outcome leads supporting the how, and the product outcome leads supporting the technical how. Importantly, we're looking for these to be at all levels, including team and value stream.

We also wanted to get crisp on how we measure improving ways of working, and at its core is value. Again, how that nested laddering occurs from team up to the strategy, and the important thing here is to use both leading and lagging indicators. We created these four quadrants of quality, time to value, team engagement, and safety. What's important to note about this is balance. For example, we didn't want to have speed at the expense of quality.

So, Anna, I think that draws to a conclusion chapter one. Where are we heading next?

Anna Steel: Thanks, Marcus. That leadership off-site in June was a really big moment for us. The takeaway was that we needed to go all-in on nimble ways of working, thankfully not in the context of big-bang change, but rather with renewed leadership support, organized and intentional value streams that would help us deliver on our multi-year objectives.

As we turn the page to chapter two, focus will be key, and nimble ways of working will be our how. This summer, we began to organize for outcomes. For the first time, we're bringing work to the people instead of people to the work. We're looking for continuous value delivery and seedling teams that are now starting to evolve into saplings. We're getting crisp on our outcomes and key results, and also starting to optimize end-to-end delivery governance through lean portfolio management.

Gratifyingly, we're also now seeing leaders lean in now more than ever, and that's all helping us make that shift from early adopter to early majority. As a result, we're starting to see more pull for ways of working, and it's been really exciting to see these efforts start to pay off. In our H1 ways-of-working engagement survey, we saw a 42 basis-point lift in overall ways-of-working eNPS. For those of you who are paying attention to the numbers, yes, overall eNPS is still negative. However, we have seen improvement across every major area in our ways-of-working survey, and that's been really reassuring. Also, a first in my 10 years at Airbnb: we've created new platforms and mechanisms to drive engagement and learning in teams, and as a result have expanded our ways-of-working footprint to 26% of global CS.

Marcus Braszell: Well, it's a great story, and I particularly like the way we've overcome some of those impediments. But don't just take our word for it, folks. Let's hear from the people who matter most, the teams in Community Support.

Team Testimonials

More of these coordinated efforts between teams where everyone within those meetings has a voice.

I like that we're shifting our focus to the outcomes we want to drive for our community, and most importantly that we're empowering CST members to help figure out how we're most likely to achieve those.

New ways of working has empowered us to deliver at a higher speed because now we have a diverse and full team where everyone complements each other.

What's most important, we finally started to focus on the right things, and that's just...

From the time we identify which project we're going to take on to doing a root cause analysis, to getting help from analytics, to help size, to move towards experimentation, it's really, really sped up.

Having multidisciplinary teams of early adopters aligned to a business outcome and with a bias to action created the social proof.

I like how we started acknowledging that we just shouldn't try to do more or do things in a faster way, but actually trying to focus on the outcome and in the impacts that the teams are generating, which is really giving the teams the autonomy to make decisions about improvements that are going to improve our agents' experience as well as our customers' experience.

I often hear new ways of driving conversations from senior leaders to ensure we are considering our own work in a more visual way and calling on blockers or impediments.

Number one, I think we're more focused. It's always needed if you want to do impactful work.

I love the energy and passion about the work.

The focus on the engagement to deliver value that comes along with working with a dedicated team has been amazing.

While we have a ton of work left to do, we're starting to see that we're able to move faster, and we're starting to move faster together, and that's really important. Changing ways of working is really hard. The team has done an incredible job guiding us through every single step of the way, starting small, delivering big.

Thank you. Collaboration is at the heart of what we do, where everyone can come to the table with their ideas, regardless of the team that they work in.

Great minds, different perspectives, and insights truly gather together to achieve ideal states, optimizations, and really ambitious goals.

Marcus Braszell and Anna Steel

Marcus Braszell: Big thank you to our early adopters at Community Support. Where are we heading next? Now you can see the full journey start to open up, and our aspiration is to work towards a generative learning organization.

The first question mark there represents our emergent mindset improving ways of working. There's no crystal ball. It's unknowable. It requires a probe-sense-respond approach. However, we've been guided by early experiments, and we have confidence moving forward.

We're also aiming for a generative organization. It was fantastic to hear Ron Westrum talk about that yesterday, and the point here is mission identity with the purpose of the organization. We're also aiming for a learning organization using data and insights, experiments to learn, and also where it's safe to fail. Finally, most importantly, a psychologically safe environment where people can do their best work and where work is humane.

Anna, would you like to recap chapter one?

Anna Steel: Thanks, Marcus. Quickly, these are the patterns that resonated with us when we thought about our chapter one. It starts with outcomes over outputs and an approach that's not one size fits all. We need to think big, start small, learn fast, and invite over inflict. That approach helped us find the patterns that work best in our context. Impediments are the path. They illuminate the dependencies and challenges we needed to overcome to chart that course ahead. To navigate, we needed to be laser-focused on our outcomes, organizing in intentional value streams that are optimized for the fast flow of measurable value for our community.

Thank you again so much. It's been a privilege sharing our story. We hope these resonate with you. Ways-of-working journey being a journey of years, we know we've much yet to learn and would love your help.

Marcus Braszell: Thank you. We'd love to continue the conversation, so in the break, please reach out to us about any of these topics. Anna, there also might be something else we'd want to talk about.

Anna Steel: One last thing. We also love talking about dogs, so if nothing else, we'd love to hear from the other Frenchie owners in the audience. With that, huge thank you, Las Vegas. We look forward to keeping you updated on our chapter two in the future. Thank you.