Cold Start Problem — Book Summary (Part II)
Just as a recap, the first step to building a new product with network effects, is to build a single, tiny network that’s self-sustaining. Countless failed products are a constant reminder of how hard it is to build such a network. Let’s discuss one such product, Glitch. It was a multiplayer game in the browser where users walked around the game world, clicking objects in the space to learn skills like gardening, teleportation, etc. The game was described as painfully boring and mostly stupid. Users who signed up, 97% of them left within 5 minutes. This leaky bucket never filled enough to unlock the social experience that multiplayer games are supposed to offer. Why is building that first network so hard, you may ask? It is because small, sub-scale networks naturally want to self-destruct. After all, when people show up to a product and none of their friends or coworkers are using it, they will naturally leave.
Naturally, the next question then becomes, how many users does your network need before the product experience becomes good? The way to answer this is by analyzing the size of the network plotted against a set of important engagement metrics. As a business, you’d want to maximize this number at least up to the point of diminishing returns. With enough analysis, there will emerge patterns that highlight the network density that is needed to spike up usage.
The higher this threshold, the higher it is to get started but the more defensible your product is in the long run.
The common thread in the launch of products with network effects is that they start small, a single city, college campus, or beta tests at individual companies. As Chris Dixon says, the next big thing will start out looking like a toy. Disruptive technologies are dismissed as toys because when they are first launched they undershoot user needs. Success lies in adding a small group of “right” people, at the same time, who use the product in the “right” way. Once you create one stable network that is engaged and can self-sustain, you can copy-paste it many many times to build a huge interconnected network that spans the entire market. Credit cards are a great example, outside the tech industry.
Credit cards aggregate consumers, merchants, and other financial institutions as a multi-sided network. Everyone in the network benefits, especially the consumer. And the bigger the network gets — meaning more consumers, more places start accepting credit cards, making the network more useful. This in turn drives new merchants and consumers to adopt it. It was Bank of America that invented credit cards and launched them in Fresno, California. The city was chosen not only because of its size and population but also because 45% of Fresno’s families did some business with the bank. They mailed 60,000 residents a credit card without an application. The fee for merchants was 6 percent and consumers received $300 to $500 in instant credit. Because of this strategy, on day 1 cardholders simply existed, causing merchants to sign up.
To pick your atomic network, first, you need to have a hypothesis about what it may look like. The size, key players, etc. The networked product should be launched in its simplest form so that the value proposition is dead simple to the target group. The attitude whilst executing is to get momentum by doing whatever it takes. Uber initially focused on narrow, ephemeral moments, like 5 pm at the Caltrain station at 5th and King St. The driver operations team used an internal tool called Starcraft to direct drivers in real-time. Doesn’t sound scalable at all. After this atomic network was established, the group moved to a wider network.
New networks should be focused on groupings of a handful of people, with the right intent, in the right situation at the right time.
Network products often have sides, buyers or sellers or content creators and consumers. One side of the network is easier to attract but the other side which ends up doing most of the work is harder to acquire. The minority of users that create disappropriate value, therefore, have disproportionate power. For example, most of Wikipedia was written by a tiny percentage of prolific editors. To attract and retain the hard side you need to design a product that is sufficiently compelling to that subset of the network. This is because they will try competitive products and their expectations are higher. A social network can’t exist without its content creators and a marketplace can’t exist without sellers.
The easy vs the hard side
The question that comes to mind is, why is there a hard side at all? The hard side exists because there are tasks in any networked product that just require more work, whether that’s selling products, organizing projects, or creating content. It is important to understand the motivations of contributing to the network, be it status benefits financial benefits, or both. The more difficult the work needed to be a part of the hard side of a network, the smaller the percentage of users who will participate. If you use a pyramid to represent the communication technology, self-expression and communication form the base. Snapchat sits there, people are very comfortable talking to their friends and expressing how they feel. The next layer is about status, showing people you are cool, etc. Now, this has a lower frequency of engagement because nobody does something cool every day. Moving to the next layer, the top of the pyramid is probably where Tik Tok lies. People who have a creative way of telling a story, or learning a hard dance are few. There is a power-law curve where 20% of the influencers end up with the most engagement.
So when building an atomic network, you have to answer the following:
Who is the hard side of the network and how will they use your product? How do they hear about the product and in what context? What is the unique value proposition to the hard side? As the network grows, why will they come back more often and engage more? The key is to find a problem where the hard side of the network is engaged but their needs are unaddressed. There are smart, motivated, early adopters who are on the lookout for opportunities to make themselves useful. Look at the trends around hobbies and side hustles. What people are doing on their nights and weekends represents all the under-utilised time and energy in the world which could be used to power an atomic network
Now let’s understand through an example, how Tinder solved the problem of building the atomic network catering to the hard side.
The key is to build a product that solves an important need for the hard side. For online dating, the problem was too many love letters, causing women to be overwhelmed with a large number of messages. This cascaded into men getting no signal and leaving the platform. Attractive people, particularly women, are the hard side of the online dating network. Tinder understood the problem and made dating fun through easy signup, making it a visual first platform and introducing the swipe functionality. They also added Facebook to connect users which mimicked the past where people were introduced to potential mates by their friends. This also meant not being exposed as a suitor to your friends. This created a lot of trust in the platform. The best part that drew women was how a user didn’t have to share their number and unmatching someone was so easy. This created an environment of safety bringing the critical mass of supply to the platform.
When the cold start problem is solved, a product can consistently create Magic moments.
You can read the first part here.