Andrew Chen, a partner at Silicon Valley venture capital giant Anderson Holtz** and a former vice president of growth at Uber, has published several articles on product, growth, marketing, and network effects on his blog, which have been featured in Wired, Wall Street**, and The New York Times.
Recently, his new book, Cold Start, was launched, which answers all kinds of questions about "user growth" and provides some ideas for startups to achieve user growth from scratch.
Cold Start [US] Andrew Chen.
People in the industry often think that dating apps are very difficult to start, and even harder to achieve scale growth, and there are a lot of cold starts in this category of software, but Tinder is an extreme exception.
After two people are successfully matched on dating apps, they will leave the platform, so the more successful the date, the higher the churn rate of the platform.
So, how does Tinder successfully replicate?
Tinder's team plans a birthday party at the University of Southern California, a party like no other. There is a small threshold to attend this party: you must be on your phone** tinder and registered as a user to attend the party. They hired someone at the door to check if the participants had done this.
The party was great and the tactics worked, there were some very charismatic people at the party who didn't have time to talk last night, and now they had a chance to communicate again, and the tactics of releasing software at the college party worked.
For Tinder's development team, the party was the highest in a single day, and perhaps looking back on it now, it wasn't a big deal. In this tactic, it's not just the total number of users that is important, but the acquisition of "500 suitable users", which is the words of Tinder founder Ladd.
The initial "seed" users who joined Tinder were the most social and networked on the USC campus, and they all joined Tinder at the same time, and these 500 suitable users were enough to influence the entire USC user base.
Tinder found a model for growth, and they went on to host Valentine's Day Tinder parties, cocktail Tinder parties, sorority Tinder parties, and all sorts of other themed parties, and each one worked as expected.
One of the key things about getting started with Tinder was discovering a strategy that could be replicated. There are many replicable user growth strategies, but there are three that are critical.
The following text is excerpted from CITIC Publishing Group's new book "Cold Start".
By Andrew Chen
1. By invitation only.
Sorry, only invited users can register as users of this software. "Anyone would be upset to see this kind of tip. This may seem like a counter-launch to a product release, especially during the cold start phase where new users are urgently needed.
Why do you turn away someone who wants to try your product?
Some people think that the "invite-only" strategy is a hype method, because a hot new product might get people to send invitations to friends on social**. Others argue that the intent of an invite-only strategy is to limit the rate of user growth so that development teams have time to patch system bugs and expand the product infrastructure so that the product can be fully functional when it actually comes out.
Both of these arguments have their own merits, but they ignore the most important reason why online products are invite-based.
Invitations are inherently copy-and-paste – you start with a hand-picked network, send out invitations to the first group of web users, and the network starts to copy automatically.
This is what happened on LinkedIn.
Hoffman has the following introduction to LinkedIn's early network structure: People at the level of Bill Gates must be at the top of the workplace hierarchy, and there will definitely be a large number of people who want to get to know him through referrals, and the number of such requests is beyond his control, and anyone who knows Bill Gates will encounter referral requests.
At the time of LinkedIn's launch, the software was certainly pointless for people like Bill Gates. However, there are still successful people in the middle of society, who are still building personalities and promoting their brands. If they hear that someone wants to get to know them, they usually say yes. This middle layer is where LinkedIn really comes into play. In order to seed the initial network in the middle of this workplace hierarchy, LinkedIn set up an invite-only mechanism from the beginning.
Their invitations are more targeted to people within a specific occupational level, but this alone is not enough, the positioning of the product must also be accurate. According to Hoffman, an important factor in LinkedIn's success is that LinkedIn has never explicitly identified itself as a "job search" software.
This really reassures a concern of users, because on this network, the information you post may be seen by colleagues and bosses, and there must be some scruples about using it. So for LinkedIn, it's safer to avoid being labeled as a "job search software." As a result, LinkedIn has adopted a more flexible positioning, that is, to become a web service software for professionals.
You can create your own homepage and connect with other people, and of course, you can search the web for new job opportunities, but that's just one of the many features of this product. This also means that when a person successfully joins LinkedIn through an invitation, they are more likely to register as a user, and they are more likely to invite others to join the network. LinkedIn exploded within a week of its launch, thanks entirely to invitation-only benefits.
LinkedIn didn't stick to the invitation system for too long. In the second week after launch, the core network user base was already strong enough that the operations team made the decision to open registration.
What started out as a core network of well-connected, aspiring Silicon Valley entrepreneurs and investors made a big difference. They built momentum in the market, attracting a group of even more important users, a broader user group that Hoffman called LinkedIn's "loyal followers." While tech practitioners in the San Francisco Bay Area continue to join the network, true believers have begun to actively interact on the platform, and their numbers are growing exponentially in all corners of the globe.
A few weeks after launch, LinkedIn has reached a tipping point — a product that fosters user engagement and brings value to groups outside of tech users.
2. Come for the tool, stop because of the network.
"Coming for the tool, stopping for the network" is one of many well-known strategies for launching network products and driving network growth at scale.
The concept of this strategy is to first achieve early user accumulation through single-user operational tools, and then, over time, direct them to engage in activities in the network. Tools can help entrepreneurs reach the initial tipping point with the required number of users. The network creates long-term value for users and can be a moat for the company.
In addition to sharing software, there are many different types of software in different industries that have adopted similar strategies: Google Office can be used as stand-alone software to help workers create documents, data sheets, and presentations, and it also provides web features around collaborative editing and commentary.
Transitioning from a tool to a network is a unique strategy, but not all networks are built that way. Tinder is unlikely to have a single-user model. By analyzing the development strategies of software such as **wall, youtube, Google Suite, and LinkedIn, you will gradually see the patterns behind them. Each product provides users with a set of tools around the themes of content editing and hosting, whether it's a profile, a resume, or a work document. This set of tools is combined with a network that allows users to interact with content, which in turn translates into user-to-user interactions.
Creating a "tool + network" combo can certainly be very powerful, but this strategy doesn't always work. It's hard to drive users from tools to networks. Because users need to change their habits, only a small percentage of users are willing to make the transition, and the development team has to think about how to keep users after the transition. A lot of users will stay at the stage of using the tool.
At one extreme, the tool is separated from the network, and the developer is simply forcing a popular tool with a completely unrelated web product. This makes it difficult to implement a "tool + network" strategy, as the number of users transitioning from tool to network is certainly small.
At the other end of the spectrum, tools and networks are inseparable, such as Dropbox's file-sharing capabilities, which is at the heart of the network. This level of integration of tools and networks can make users feel that it is unreasonable to separate the two. Users of such a product are more likely to drive the transformation of the product to the network than the other way around. This tool-to-web conversion rate tends to be high.
3. Pay for product launches.
One of the most common voices questioning fast-growing startups is, "Is this business going to make money?" ”
That's the kind of skepticism that Uber has been subjected to. Before going public, Uber burned billions of dollars every year, despite the company's rapid growth. Since its founding, Amazon has also suffered 17 consecutive quarters of losses. Getting a network on its feet is very expensive and costly.
For online products, especially in the early days, it makes sense to spend a little more money (and spend it like crazy) in exchange for growth. The purpose of spending money is to get the market for products to reach a tipping point as soon as possible, so that the network effect can have a positive driving force, and then the subsidy policy will be tightened. If this strategy is executed properly, the end result will inevitably be a product with fast growth and more money.
In the field of shared mobility, there are also opposing issues. When launching a new market in a new city that requires both drivers and passengers, which side would you launch first? The hard side is the default preferred starting side.
Uber has subsidized the driver community from the beginning. They post jobs on Craigslist, promising drivers a guaranteed $30 per hour for a guaranteed income, whether they receive an order or not. All the driver needs to do is make sure the software is on all the time. Solving the cold start problem with an hourly minimum wage is expensive, but effective. This burns money too quickly, and unfortunately, as the market grows and more drivers are needed, the method of paying to hire people is ultimately unsustainable.
To combat this, Uber's operations must shift to a "commission" model, which means that the guaranteed income subsidy is discontinued and returned to the regular business model, where drivers receive a percentage of the fare.
To encourage this transformation, company executives have created an internal leaderboard that provides a visual indication of how quickly operations teams are driving this transformation. Every city needs to quickly launch a new market, first pay drivers a guaranteed hourly wage, pool a sufficient number of drivers and passengers, and then transition to a sustainable payment model.
Healthy competition between teams everywhere is making this business model work faster and faster. That's where tipping points come in. The next step in expanding the initial user base is to attract more drivers. That's when the power of a network of users is most effective through a user referral program. Uber's driver app prompts drivers to join a user referral program ("A friend gives you $200 back when they sign up as a driver") to further leverage their purchases on Craigslist.
Of course, the convenience and income level of the Uber platform spreads purely through word of mouth, as drivers tell their friends. The user referral program and word-of-mouth approach have brought nearly 2 to 3 drivers to Uber.
Fueling growth with direct payments may seem like a dangerous move that should only be used under the right circumstances.
Selectively not making a profit is also a smart way to push the network beyond the tipping point. When you're able to create a few atomic-scale networks, you may want to pave the way with money and capture the entire market as quickly as possible. For marketplace networks, buyers need low prices and sellers need high profits, both of which have their own value appeals. The same is true for social products, whether it is communication or content sharing software, content creators need to accumulate a sufficient number of readers (or viewers) and be able to make money.
In these networks, the start-up strategy release may need to spend more money as a pre-subsidy, whether it is to use the money to buy early content, or to use the money to pay the seller in the marketplace a guaranteed income. Once scaled, network operators will have to reduce the use of financial incentives because they have already won the market.
A seemingly unprofitable situation in the short term may be exchanged for a long-term market dominance, provided that your market reaches a tipping point in your favor.
From 0 to 1, teach you step by step.
Build a new framework for the growth of network products.
Cold Start