13 Metrics to Watch If Your Business Depends on Network Effects (Part 2)

In our previous piece, we reviewed the fundamentals of using a data-driven approach to growing a network effects-based business model. We also looked at the two basic types of network effects-based business models: content-first and connection-first.

What Type of Network Do I Have, or Do I Want to Build?

What if you aren’t starting from scratch – what if you’re working on an existing product in an established market – how do you know which of these models applies to you most?

The answer lies in understanding what your network can’t do without: is it content, or interaction? If removing content from your network breaks the network effect, then your network is based on a content-first model. Similarly, if removing the interaction around content from your network breaks the network effect, then you’re likely to have a connection-first model.

It isn’t always a clear or simple distinction – but most existing networks will display a preponderance toward one dynamic over the other. Understanding which is the dominant one for your own network will help you determine which metrics to use, how and why.

Building a Content-First Network Effects-Based Business

The focus in a content-first network is on quality: both in terms of the content, but also the types of interaction around the content that is created as well, for it is this interaction which underlies the growth and health of the network effect.

This translates to an intense focus on the standard of content you include in your business. Also, it means broadening your understanding of what content actually is, as it can also include the interactions which flow from it: comments, shares, or inquiries. In fact, any interaction which drives public – or at least shared – participation can be regarded as content.

Famous Examples:

Twitter, Facebook, Reddit.

Visitor Engagement:

Engagement is one of the most foundational measures of content-first network. Fundamentally, you should be aiming to convert occasional users into regular users through the quality of your content.

The most effective way to measure: Try measuring this using a day-to-week ratio: how many of today’s visitors were here earlier in the week? Alternatively, try measuring the number of days since a previous visit for returning users.

Content Creation and Interaction:

The key split here is between content creation and consumption. Depending on the business or industry you are in, you may find that user behavior tends toward one or the other. Certainly, the way you build your network will typically lead users toward one behavior over the other.

The most effective way to measure: The ratio between creators and consumers is a good point to begin, though not in isolation. It would be a mistake to assume that you should aim for balance in numbers for both types. Instead, you will need to reference additional external behavior. Try measuring the rate of content creation initially, and then begin to focus on the quality of that content by looking at the amount of comments, likes or shares associated with it, treating these as indicators of quality.

Engagement Funnel Changes:

Understand the funnel which represents engagement by your users for your site. The number of tiers or stages involved in how users interact with your network will in turn dictate how you can optimize their experience of your site. Not all engagement is as simple as tracking the time a user spends on a site: instead, truly understanding engagement means being able to relate specific individual interactions back to behavior that causes a user to convert to the next – and higher – level of engagement.

The most effective way to measure: Try comparing different tiers of engagement over time. Given one segment of users, how long does it take them to convert to the next level of engagement, and is the level of conversion high enough? In this way, you can begin to focus on specific interactions which exist at the transition points in the engagement funnel.

Value of Created Content:

Value can be relative for content-first networks. Of course there is the economical business value that content generates, but this is only generated by understanding the multiple levels of engagement that users have with content.

The most effective way to measure: Track this as a trend, referencing its behavior over time as a good way to understand the underlying value that is being generated. Measure this by traffic segment or cohort, so that you can track the value relevant to a specific improvement you have made or a campaign that you have initiated.

Content Sharing and Virality:

Measuring virality and sharing as distinct behaviors gives you an additional level of detail to understand how your content can explicitly be a source of growth for your network.

The most effective way to measure: Understand the dynamics of virality in order to know what your “critical mass” of sharing will need to be. Also focus on tracking visitors in as much details as possible.

Notification Effectiveness:

Notifications can take many forms, and the way you measure and monitor them depends entirely on the medium through which users will interact with your network. If your network is based on access to a native application on a smartphone or tablet, then notifications will exist outside of the context of your product. Alternatively, if your network is interacted with through a Web-based application, then you may have less control over the way those notifications are delivered.

The most effective way to measure: Measure notification effectiveness in the same way that you might measure delivery rates for an email campaign. You want to be able to assess the efficacy of these notifications in driving users back to your network, whatever type of application it is.

Building a Connection-First Network Effects-Based Business

In many ways, a connection-first network is like a two-sided marketplace. These are characterized by networks that contain two types of users: buyers and sellers. Conceptually, if not literally.

What defines a connection-first network is that user value (and thus business value) are derived from the interaction between both user types on the network. Dating sites are a great example: the desire for interaction between users is what spurs content creation (in this instance galleries, and profiles).

The chief challenge of connection-first network models is that they must sensibly identify the segments of each user type they need to reach, and also grow the numbers of each type if they are to successfully grow their network.

Famous Examples:

Kickstarter, Indiegogo, Pozible, Tinder.

Growth Rate of Users (Members/Time)

As mentioned in the first part of this article, a good metric is usually defined by its ability to predict what will happen, and help you sensibly determine what you should do next. Rates are the epitome of this, comparing multiple numbers against each other to help you understand what’s happening.

The growth rate for users in your network is one of the key data points that should guide your behavior. The principal to achieve is consistency and balance: growth and decline are managed by focusing on the addition of one type of user over another. As with the dating example above, if a dating site has too many men, then network growth will eventually stall until the imbalance is addressed.

The most effective way to measure: Try focusing on this mostly during the early stages of network growth. Manage the trend, not the number.

Rate of “Inventory” Growth (“Listings”/Time)

In addition to membership, you also need to track the trends behind the key behaviors that users exhibit in order to keep them coming back. In the example of eBay, you’d call these listings. It’s not always this cut-and-dry, though. Being a connection-first network requires you to identify which behaviors act as inventory: for example, managing the number of active invites in a network could produce a similar result.

The most effective way to measure: In addition to tracking the rate of growth, also try to track completion of the behavior as well.

“Buyer” Searches (The Number of Searches That Return 0 Results)

Often, search is the main behavior and activity that drives connection in a network. Be it finding friends you know (Facebook) or connecting with people you may not know well at all (LinkedIn), the main thing to track are searches that yield no results at all. Tracking this allows you to identify any unmet needs that users have.

The most effective way to measure: Once you are tracking searches that yield no results, your main objective should be to address the reason for them occurring – learn as much as possible about what these unmet needs are telling you about your users, and how you might best address it.

Conversion Rates and Segmentation

Understanding the way people flow through your network is key to understanding how you can influence their behavior. A distinguishing characteristic of networks is that the user journeys through them are rarely linear – users will often repeat tasks multiple times. Also, the behavior that allows you to segment them into distinctly serviceable group is usually in reference to other users of the network itself.

Understanding these underlying dynamics will help you to identify the stages of user experience in the network, and then deliberately manage the transition between them.

The most effective way to measure: Segment your users by their main self-identified differences, either through interviews or by direct observation of their behavior. Then, identify the primary and secondary ways that they pass through your network. Begin by optimizing your network for the primary user flows, and then extend your efforts to any secondary or tertiary user flows.

Buyer and Seller Ratings

A key characteristic of a connection-first network is that sentiment must be measured and managed explicitly in order to ensure a quality interaction between both user types on the network. Without doing so, maintaining positive network growth is hard to do.

The most effective way to measure: This often takes the form of “ratings” in many platforms, though you may develop your own form of sentiment tracking. The key is to identify how sentiment can be tracked in your network, and aim to have it continually trending upward.

Related reading