The Cold Start Problem 3/7
8/ Solve a Hard Problem
Attracting a hard side requires building a product that solves an important need for it. (e.g. Tinder: the hard side are attractive women.) Too much low-quality engagement is not better than the insufficient amount of engagement.
For marketplaces the hard side are the sellers (the supply side) and the value proposition to many small sellers is the ability to have a side job. Bringing initial supply has higher priority than demand – but the latter should follow soon. And then the focus will shift to supply again, for much longer.
The opportunity for startups lies in finding a problem where the hard side of a network is engaged, but their needs are unaddressed. (a direct quote) Look at hobbies and side hustles. There’s no shortage of people who can apply their skills and contribute their knowledge and interests in order to try a new platform. Nights and weekends are an untapped time resource for creation.
Spare time (social networks: content) and underutilised assets (Uber: cars; Airbnb: rooms; Craiglist: old stuff to be sold) can form a basis for attracting the hard side.
The hard side should also be segmented to find out the underserved subsegments (since everyone can’t be equally happy) and provide them with additional functionality for creation / distribution / monetisation. For underutilised resources the logic is similar: helping the hard side find a more profitable and/or frequent use of their assets improves retention.
The traditional approach of going bottom-up (start with a cheap segment and go upscale), which is considered a de-facto way of growing a business, is actually not the only one: making cheaper versions of a luxury product may not dilute the brand positioning of this luxury brand (case in point: Uber that started with black cars and added cheaper segments later).
Atomic networks are usually built at the low end of functionality in niche markets; later their hard side demands access to other verticals and higher-end opposite user segments. (Case in point: Airbnb started with mattresses, then rooms, then apartments, then castles and exotic dwellings.)
9/ The Killer Product
Sometimes (as in the simplified case of Zoom) all that’s needed for a new product to take off is the ease of use (vs competitors) and higher quality of the core functionality (video in case of Zoom). Frictionless user onboarding improves the network effects.
The fundamental difference between the networked products and the traditional products is that the core functionality of the former revolves around facilitating user interactions, and of the latter – around facilitating the interactions between users and software. Thus, the networked products benefit from having more users, while the traditional products benefit from having more features (some of which are barely used). [MK: this is not to say that it’s an “either-or” type of choice.]
Discovery (be it content or another user) is an essential part of networked products.
Simplicity of use is a good driver for increasing the usage frequency, itself leading to introducing more users to the product.
Focusing on users instead of features makes networked products look deceivingly simplistic and even barely defensible. [MK: to state the obvious, different users have different mental capabilities for interacting with an app, so it’s better to keep the UI as simple as possible. It’s this old “don’t make me think” principle in action.] Consumerisation of enterprise applications is actually a good idea if the goal is winning the hearts of employees and broadening the user base within an organisation.
Defensibility lies in the retained user base, which can’t really be stolen even if competitors get access to every individual from it.
Free (or freemium) is a preferred distribution model for networked products as it eliminates the barrier to adoption and forms a habit before the user is asked to pay a single penny. Video and social media platforms are usually ad-supported, so users get monetised instantly (or can pay to remove ads, see YouTube Premium), workplace/B2B services get monetised via subscriptions; games, marketplaces, livestreaming platforms get monetised via micro transactions. However, asking a new customer to pay even a penny substantially reduces the appeal and taints their onboarding process.
New technological features (GPS, good cameras, high-speed internet, fast payments) gave rise to lots of unicorns using them in one way or another. Shift in technological platforms’ capabilities gives rise to new products that have a potential to dominate the market.
While it sounds cliché, the massive shift from desktop to mobile has created new demand for products and developers almost overnight.
Once the killer product and the first atomic networks are built, it’s time to create the “magic moments” :)
10/ Magic Moments
Magic moments occur when the product works as intended, users enjoy its value and there’s enough content for everyone. But how would one measure them?
Start with “zero [MK: value]” – a moment in the user experience with the worst possible experience (assuming a non-crashing product). There may not be enough users on any side (easy or hard), or no content (cars, listings), or no reason to interact with a product. If there’s a certain task in mind and it can’t be fulfilled by the product – that’s zero value.
Getting from zero is hard, but it also requires maintaining this “above zero” experience. Experiencing “zero” in the course of using a product (having a request unfulfilled) increases churn and reduces trust in the product; at some point churn can exceed the active user inflow and the product will start dying.
“Zero” is a very important aspect of user experience and retention that companies have some control over; instances of “zero” need to be recorded, monitored, and dealt with. [MK: I guess the #1 predictor of churn is a row of “zeros” for a user.] While it may not be practical to keep the percentage of users who receive “zeros” at, well, zero, every product has its threshold exceeding which causes the user base to irreversibly shrink.
Answering the question “how would one measure magic moments?” – it’s done via observing the growth of the key metrics: acquisition via brand queries, user invitations and content sharing, engagement with the content or with each other, increasing user retention, better monetisation and the emergence of new customer segments.
Growth in new segments has its own Cold Start Problem, as initially the network will stretch itself thin, and it will require time and/or conscious effort by the product team to make this new network dense enough.
Part 4.