Former Airbnb Chief Growth Officer: How to Build, Refine, and Scale a Growth Program?
TechFlow Selected TechFlow Selected
Former Airbnb Chief Growth Officer: How to Build, Refine, and Scale a Growth Program?
Growth is crucial for startups because it's in their very nature.
Navigating Web3 tides with focused insightsCompiled by TechFlow
Note: This article is part of the TechFlow series "YC Startup School Chinese Notes" (updated daily), dedicated to collecting and organizing Chinese translations of YC courses. This is the seventh installment—YC partner Gustaf Alströmer’s online lecture titled "How to Get Users and Grow."
Why Growth Matters?
Before joining YC, I was Airbnb’s Head of Growth, where I saw the team grow from three people to around 120. Almost everything I’ll discuss today comes from that experience, as I learned a great deal about best practices in user growth.
Growth is crucial for startups because it's essentially what they are all about.
Paul Graham’s essay “The Essence of Startups Is Growth” explains this in detail. However, applying these skills and insights requires certain preconditions.
First, you must have product-market fit.
Second, you should be working at a consumer or online company, as these types of companies typically adopt such best practices.
However, if you try to apply these methods before achieving product-market fit, it could lead to poor outcomes. Therefore, ensure your product meets market needs before deploying these strategies.
Facebook’s Growth Journey
In its early days, Facebook already had a strong data science team skilled in measuring and forecasting its scale. Around 2006, shortly after forming a growth team, Facebook’s data scientists began predicting user numbers, estimating based on available data that Facebook would reach about 400 million users by 2015.
But in reality, Facebook experienced rapid growth starting around 2008. Initially available only in English, Facebook later introduced translation features, which re-accelerated growth.
Around 2010, increasing numbers of users accessed Facebook via mobile devices. The rise of mobile technology posed major challenges, forcing Facebook to adapt and use data to solve them. They overhauled their entire team and even launched extensive training programs so engineers could learn mobile development.
By 2013, Facebook had reached peak internet penetration with massive active usage. To sustain growth, they launched "internet.org"—a project aimed at connecting more people to the internet. Partnering with airlines and telecom providers, they offered free access to Facebook. This became a key component of Facebook’s continued expansion.
Measuring Retention
If you want to grow rapidly and surpass predictions and limits, you must actively drive growth—and this applies to every company.
However, having a growth team before product-market fit isn’t truly useful.
One way to assess product-market fit is to identify standard metrics and measure how frequently users engage with them.
For example, for Airbnb, the metric might be repeat bookings; for Facebook, it could be how often active users return.
If you can answer these two questions, you can perform cohort analysis on your own company to determine whether you’ve achieved product-market fit.
Ultimately, by measuring how frequently events occur, you can chart whether your company has achieved product-market fit.
Although Shopify has many registered users, the number continuing to register in the second month alone doesn’t indicate product quality—some will naturally stick around.
If the retention curve drops to zero or keeps declining, then the product likely fails in the market. If product-market fit is strong, the focus should shift from searching for fit to driving growth. In such cases, companies should aggressively scale and implement measures to retain users—even if some are initially lost.
For instance, if 50% of users remain after one month and 10% after 24 months—similar to shopping apps—users are still being lost. Even if users start paying after 12 months, their monthly spending may gradually decline. So it's hard to claim full product-market fit, as they haven’t yet found the optimal strategy.
Netflix, on the other hand, retains 70% of users who pay and use the product for over seven years—a clear sign of strong product-market fit. Such companies should do everything possible to fuel economic growth.
Retention rate is one of the most important indicators of product success and can be measured in various ways—for example, through user interviews asking whether they’d continue using the product.
Additionally, it’s essential to know whether your current product offers real retention value. If not, pursuing growth is wasteful and drains time and resources.
Historically, product and marketing teams were separate. Today, diverse roles—including effective product managers, engineers, data scientists, and marketers—can all drive growth. These teams leverage technology and data to modify products and attract more users.
Brand marketing is one of the hardest types of marketing to measure. Startups shouldn’t invest in brand marketing until they’ve reached a certain scale.
Funnels are critical components of any product, consisting of multiple steps. Growth teams improve product success by optimizing conversion rates. Key stages like authentication, registration, onboarding, and purchase conversion are particularly challenging and require continuous optimization. For international rollouts, translation helps more people use the product. Authentication and registration also demand high attention. Finally, purchase conversion is another vital success metric requiring targeted actions to guide user decisions.
There’s something called a “growth channel.” What exactly is a growth channel? It’s how people discover your product.
When you’re a small company—if you have fewer than 50 users—you shouldn’t think about growth channels. Even with under 5,000 users, it might still be too early.
But when you're small, if your activities don't scale, problems arise. Most successful companies grow through scalable growth channels.
These growth channels aren’t generic platforms—they must be used within specific contexts. Few channels match Google’s breadth, but many other products are being discovered worldwide.
First, consider how people find your product—what rare behaviors do they use to discover your solution? For example, when buying a home, people may search via search engines.
Once users engage daily with your product, they no longer need to return to search engines—they can open the app or visit the site directly. Thus, when optimizing growth channels, focus especially on SEO and paid search.
In short, Google isn’t the only search engine, and others can be optimized too—but in most cases, Google remains one of the most important.
Growth Channels and Strategies
Expanding product reach by accelerating existing user behavior is a viable approach. You can incentivize referrals through monetary rewards or offer free methods.
Does having more users genuinely enhance my experience?
I mean, if I’m building the next LinkedIn, the product becomes less valuable once everyone uses it.
But if more individuals and companies begin using LinkedIn, the added value becomes meaningful.
From this perspective, you should aggressively pursue growth, since each new user presents an opportunity to attract more. This is a proactive mindset.
Now, here’s a common question: Can you list everyone who uses your product—literally, like a spreadsheet? Say I sell to people visiting doctors. Well, I could list all doctor offices in the U.S. or California—that’s not difficult.
So I’d find a way to build that list and then do sales. This is an amazing fact—often, this is exactly where you should start.
Lastly, do my users have high LTV (lifetime value)? Is my product expensive enough? If so, I should definitely explore acquisition channels like Google and Facebook. But I should only spend on ads or acquisitions when my product delivers real value and is accepted by customers.
I worked extensively on Airbnb’s referral program. Referrals are somewhat like a well-designed monthly plan. If people are already talking about your product, referrals amplify that conversation.
Financial Incentives
One way to simplify things is through financial incentives.
In Airbnb’s referral program, we implemented an economic incentive: as a referrer, I earned $20 for each user who signed up with a travel credit, while the referee received $40 in Airbnb travel credit. We started with this principle and maximized our referral channel to reach as many people as possible. Even the referred product had its own funnel—an internal Airbnb product. I won’t go into all the details, but you need to break it down systematically with an engineering and product-oriented mindset, dividing it into steps and measuring conversion rates at each stage.
For Airbnb’s referral program, we analyzed conversion rates at every funnel step: how many active users saw the referral program, how many invitations were sent, and the new user conversion rate.
We found only a small fraction of active users actually saw the referral program—if you don’t see it, how can you use it?
So we began optimizing the funnel to increase visibility and improve conversion rates. We broke the funnel into finer segments and continued refining each part.
Running Experiments
In Airbnb’s early days, we ran many experiments—such as improving referral email copy—for better results. These were carefully designed, with every step serving a purpose. Ultimately, our efforts paid off, making the referral program one of Airbnb’s key success factors.
First, email subject lines should include the sender’s name—people are more likely to open emails from friends. The subject should clearly convey value. For example, “Get $40 off your first Airbnb stay” is clear, specific, and creates urgency, prompting readers to act before the deadline.
Email content should be extremely clear—don’t make readers guess or waste time. Include social proof, like the sender’s endorsement of the product, and use exclusive language such as “Accept invitation” to prompt action.
Paid Growth
The key factors are customer acquisition cost (CAC) and user lifetime value (LTV). If CAC is lower than LTV, you have positive return timing. For example, if CAC is $20 per new user and they pay $10/month, you achieve payback in two months.
Attribution is also crucial, especially when using multiple ad platforms simultaneously. Regarding paid growth, only a few channels scale effectively: Facebook, Instagram, Google, and YouTube. However, note that organic growth channels are declining while paid channels are rising.
Search Engine Optimization (SEO)
In SEO, focus on two areas: on-page optimization and backlinks. Although some believe SEO is outdated, it remains critically important—Google is one of the world’s largest websites and shapes much of our digital future.
Keep in mind: what Google sees differs from what humans see. For example, Google cannot interpret images or JavaScript, so proper steps must be taken to ensure indexability.
For on-page optimization, start strategically. Define your ranking goals and conduct keyword research to identify all relevant search terms related to your product. Once targets are set, other on-page optimizations become easier.
Smaller companies can experiment with best practices, but large organizations must rely on experimentation for decision-making. At Airbnb, a 20-person team—with 12–13 engineers—was dedicated solely to SEO.
Off-page optimization is equally important—consider who links to your site. Use tools to identify all referring domains and their authority. High-quality press coverage significantly boosts backlinks. Without media coverage, earning links becomes difficult.
Fewer websites now link broadly, so you must strategically choose who links to you. The simplest tactic is ensuring anyone mentioning your name links back to your site.
A typical growth team includes engineers, data scientists, designers, product managers, and user researchers. There are two ways to structure such a team: as a standalone unit or integrated with the core product team. The right approach involves setting clear goals, defining boundaries, and forecasting outcomes before launching initiatives.
Using Experiments to Validate Decisions
If you ask any Airbnb product manager, “What’s the most important tool or lesson from Airbnb that you’ll apply to your next venture?” they’ll say experimentation, experimental frameworks, or the company’s A/B testing methodology. This emphasis appears even in investor decks. Most decisions made by intuition or guesswork succeed only by luck—or fail outright. Without data and experiments, you risk failure. Even at Airbnb’s scale—or smaller—if you skip A/B testing, you can’t truly understand the impact of your decisions.
At Airbnb, we experimented on every major decision across the company. Suppose your company decides to launch a new feature—you need a framework to decide. By running two versions of the same feature or site simultaneously, you can see the real difference between launching versus not launching. This is called counterfactual analysis, commonly known as A/B testing. While this may sound technical, it’s vital—because once you succeed, you assume you’re good at decisions, but they only get harder.
Internally, using experiments and A/B testing for product decisions was paramount. Without data and experimentation, decisions rely on guesses or gut feelings—often leading to failure. Experimentation is a tool for evaluating different product features and selecting the best option.
For example, Airbnb aimed to increase mobile app signups. They tested two versions of a sharing form—one increased registrations by 40%, proving the importance of experimentation. Another test compared different share button designs. Both round and square buttons performed well, showing how hard it can be to predict the right choice. Hence, experiments and A/B tests are essential for sound decision-making.
Growth Advice for Early-Stage Startups
First, measure product usage. Then pick a metric, set a goal, and run A/B tests to help make tough decisions. Relying solely on intuition or guesses leads to errors—use experiments and data instead.
Regarding experiments and SEO, test changes in organic traffic volume and ranking shifts. Since Google constantly adjusts rankings, major website changes require A/B testing. For instance, modifying title tags might boost click-through rates and improve rankings.
Whether early-stage startups should run A/B tests is debated. With low traffic, A/B tests may yield unreliable results.
Yet, with sufficient traffic, A/B testing helps validate small-to-medium changes. Use an A/B test calculator to estimate required sample size and effect magnitude to determine if testing is worthwhile. Overall, starting A/B testing as early as possible is better.
How to Approach Growth in High-Barrier Markets Like Healthcare Insurance?
Applying growth principles to high-barrier markets like healthcare insurance is challenging. I believe you must separate overall market growth from your company’s growth within that market.
- If reaching target customers is difficult, try expanding outreach to achieve some level of growth.
- If barriers stem from risk-sharing or regulation, growth principles may still apply.
- But if growth relies heavily on sales, consider automating parts of the sales process with technology.
For example, a startup selling to health insurance buyers can expand reach by emailing more people. Emails can maintain a personal tone and be addressed individually to boost open rates. Explore additional outreach methods to fuel growth.
However, remember growth can’t solve all market issues—especially those involving risk and regulation. Startups shouldn’t conflate solving systemic problems with driving growth. Growth accelerates finding more healthy insurance buyers, but shouldn’t be the sole strategy. In high-barrier markets, consider multiple factors to develop a balanced growth plan.
What Measures Ensure Sustainable Growth?
If your business isn’t scaling, you have two options:
- Stop unsustainable activities—they don’t scale;
- Or create a playbook based on learned skills and replicate it across cities.
Suppose I manually recruit Uber drivers—this may work for the first 20 drivers, but eventually, posting job ads on Facebook will yield better ROI.
Thus, doing manual work early on—like personally recruiting users—might be necessary. But this method eventually fails due to lack of scalability. You must find alternative channels to maintain high ROI.
Most companies transition from manually creating content, to engineering-driven content creation, to modifying websites for greater search traffic.
Are Incentives Important?
I believe incentives are highly scalable—just like Airbnb’s referral program.
Hundreds of thousands of users sign up daily, mostly through referrals. Providing subsidies can thus be scalable. While handing out coupons might not scale, distributing them via email, WhatsApp, Messenger, or other channels can be—so long as you don’t lose money.
But you must calculate ROI carefully and recognize this as an added expense. Without incentives, those users wouldn’t adopt your product.
Therefore, as long as you manage this well, the approach works.
How to Leverage Free and Paid User Potential for Growth?
First, consider paid user retention. Do they truly stick with the product? Next, examine free user retention. Does behavior change once they start paying? Also monitor conversion rates—how many free users turn into paying ones. There are many ways to achieve this. One method is offering limited free access, then charging afterward. Ultimately, the goal is to maximize engagement from your most valuable users.
If your top users are paying customers but stop using your product after three months, you face a deeper issue than just managing free vs. paid users. Retention duration is critical—track usage patterns to confirm sustained engagement.
Is There an Ideal Experiment Frequency?
Sometimes running experiments is very hard—especially for market-related aspects. Setting up experiments may be far more complex than simply using online tools or configuring basic A/B tests in Mixpanel. But if you have a simple product funnel, one engineer, and substantial traffic, experimentation usually isn’t too difficult. However, some products make experimentation harder. During high-traffic phases, investing in infrastructure becomes essential for data-driven decisions—otherwise, you risk making wrong choices or spending excessive time correcting them.
In terms of frequency, experimentation has costs and should be minimized. Testing the simplest, smallest possible element is crucial. If testing a new feature, start with just its first component—not the whole thing. If you lack experiment ideas, revisit user research or study other products (e.g., Pinterest, Airbnb, Facebook). These are often highly optimized, with many funnel elements improved through A/B testing. If running a mobile app, numerous tools exist to simplify experimentation.
While experimentation can be difficult in some cases, its importance demands solutions. Investing time and effort pays off through higher conversion rates and increased revenue. Therefore, this issue must remain a priority.
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
深潮TechFlow
