Welcome back to Composed, where we provide a bite-sized piece of monetization strategy each week.

If you’re not a subscriber yet, hit the button below to keep getting these emails delivered to your inbox.

Silicon Valley loves the term "network effects" so much so that I have seen it applied to everything from online mattresses to smoothie machines. As Casper’s (lack of a) stock price can attest, you cannot simply call your business a “network” and expect it to lead to profitability.

Real network effects exist when the willingness-to-pay for a product increases when other customers use the product. Buyers and sellers in a marketplace (e.g. Uber, eBay) and interpersonal networks (Facebook, fax machines) are canonical examples.

If the value I derive is positively affected by other people using the product, then we have network effects. As a user of Facebook, I care that other people use Facebook¹. As a proud owner of a Casper mattress, I do not care whether you, dear reader, own a Casper mattress.

The strongest argument to underprice your product is when you have network effects in your market, so much so that many famous tech companies have given their products away for free, just to jump start a network.

These stories are interesting, but from a monetization strategy perspective, they are problematic. First off, we don't know whether any of the above examples monetized too late. Facebook seems to have done it well, but Uber really struggled after their IPO trying to raise prices. Some (like Glassdoor) never really monetized at all.

The more interesting case is when maximizing customers in segment A allows you to monetize segment B heavily. Facebook and Google Ads are great examples, but so is Sony. Sony's gross margin on PlayStations is minimal; they make all of their margin from allowing game developers to develop for their customers.

Ideally we want 2-sided (or multi-sided), unidirectional network effects. Let me explain:

Same-side network effects occur when a product's value increases with more customers of the same type. If 1 person in the world has a fax machine, it isn't helpful. If everyone in the world has a fax machine, it's at least a little more valuable (if people can figure out how to use it). Examples include Glassdoor (more contributors = more value), WhatsApp, and World of Warcraft.

2-sided network effects occur when a product's value increases with more customers of a different type. As an Uber rider, I don't care how many other riders there are in my neighborhood, I only care how many other cars there are. Similarly, as a driver, I don't care about the other drivers, I only care about the riders. Any network of buyers and sellers works this way (ride share, online auctions, video game consoles, etc.). Note how the value is bi-directional, e.g. riders care about drivers and drivers care about riders.

A unidirectional 2-sided network is Google Ads. As a user, of Google Search, I do not get more value from Google when other people are searching. Nor do I get more value when Google sells more ad space. But as an advertiser, I get much more value when lots of people are searching. That means that Google can make searching free and only charge advertisers without ruining the network! Another example is Robinhood. Robinhood makes $0 from users executing trades but charges market makers for the right to that order flow. Robinhood stock traders don't care if there are other stock traders on the market, and they certainly don't care about the number of market makers. But those market makers really care about the number of users on Robinhood!

Why do we care about the difference? Same-side network effects are hardest to monetize because by charging for the product, you hurt the network directly. The more I charge for fax machines, the fewer fax machines people buy, which lowers the value of a fax machine. To execute this strategy well, you have to have impeccable timing. Even WhatsApp never really got it right.

2-sided, bi-directional networks are easier, but still challenging. If I raise prices on Uber drivers, there will be fewer drivers which makes it harder to find a ride, lowering rider count as well.

Unidirectional networks are the easiest, especially if there is a same-side network effect on the part of the network you aren't monetizing. Facebook is the perfect example - by not charging for access, you maximize user count. Maximizing user count makes the platform more valuable to the next user. And maximum users looks the most attractive to advertisers. The counter example would be traditional media. There is no same side network effect for readers of the New York Times, so it wouldn't help them in any way to under charge for the paper.

Most network effects are discussed in terms of customers or segments, e.g. more riders increases value to drivers, but this same dynamic can happen at the product level. While not technically a “Network Effect” in the true definition of the word, whenever the usage of product A improves the value of product B, we have the same dynamic, and you should consider underpricing product A³.

Many of our clients are in what can broadly be described as “Vertical Marketing Technology” - software that helps businesses sell their wares. Often these are a suite of tools that might help a small business get more sales - e.g. a CRM, website, email marketing tool, and review manager. The question we always get is, “should we charge for each piece of functionality or give it away for free?”

Charging for each product is the “default” strategy (see here for what that means). That doesn’t mean you need to sell it a la carte, but it does mean you that would at least offer additional functionality in more expensive packages.

Occasionally, we recommend employing an adoption strategy by giving the additional functionality away for free. When? When the usage of the product contains minimal value in a vacuum, but greatly improves the value of the core suite.

Owner.com is a great example of this strategy. Owner’s core product is an online ordering system for restaurants, but they also offer a website builder product. You don’t have to use their website builder, but if you do, Owner promises better checkout results for online ordering. Put another way, the website increases the value of the core online ordering product. Owner should (and does) employ an adoption strategy with their website, giving it away for free, despite the fact that some customers might pay extra for it. They deliberately sacrifice revenue in exchange for higher willingness-to-pay elsewhere.

I like to bring up non-tech examples that show the wrong reason to underprice, despite having the exact business feature we’re discussing. Last week, we talked about how gyms, despite having economies of scale, shouldn’t underprice their memberships.

One service with a literal “network effect” is a speed dating. Imagine I want to gather the most eligible bachelor(ette)s in one room for a “meet and mingle”². Importantly, there will be NO SPONSORS for the event. Everything is funded by ticket sales alone.

In this business, there are obvious network effects. The more people that come to the event, the more valuable the event is for everyone. AND YET, we should not underprice for the event, because there is no way to capture the value we are creating for ticket holders (except to raise prices, which defeats the purpose of underpricing).

Instead, we would price tickets at a price which maximizes revenue. Set the price such that P*Q is highest, even if by lowering price, we would have more people at the event.

This is because the event I’ve described has same-side network effects. If instead, you had unidirectional network effects by including sponsors, it WOULD be strategically important to underprice ticket holders and make money from sponsors.

If you’ve made it this far, you should now know that the reason you can’t underprice a fax machine is because they have same-side network effects. They also have very high marginal costs. And nobody except doctors’ offices and the DMV send fax machines anymore.

Fax machines also have zero positive product externalities - I’m pretty sure using a fax machine makes every other product worse as a matter of fact.

Do you have network effects? Let's find out.

By having more customers, by how much do you think you could raise prices? At what point do you break even by lowering prices for some customers, but raising prices for others?

Crescendo works with medium-sized software companies to improve their pricing, packaging, and promotion strategies. If you’d like to book a quick consult, reach out at info@crescendo.partners or schedule time via the button below.