[This post was co-authored by me and my colleague at a16z, D’Arcy Coolican (@dcoolican). This is a companion to our other essay exploring the dynamics of network effects. If you’re an operator, we’d love to hear your thoughts on how you measure your own product’s network effects!]
Network effects are one of the most important dynamics in software and marketplace businesses. But they’re often spoken of in a binary way: either you have them, or you don’t. In practice, most companies’ network effects are much more complex, falling along a spectrum of different types and strengths. They’re also dynamic and evolve as product, users, and competition changes.
For founders, it’s important to understand the nature of your company’s network effects — including deciding on the set of metrics that help you understand what’s working or not. So, building on our previous metrics lists (here and here), we’ve compiled a list dedicated to measuring and teasing apart network effects in particular. We share them below, divided into 5 main categories to measure network effects: acquisition, competitors, engagement, marketplace, and economics-relatedmetrics.
Every single network effect business is different depending on the particular product, audience, and environment, so there’s no-one-size-fits-all list of measures. In general, however, for two-sided marketplaces matching supply and demand, pay special attention to the marketplace and unit economics sections; for social networks(including workplace ones), what matters most is engagement and activity. In the end though, it all comes down to the very definition of network effects: whether your product becomes more valuable as more people use it. Because only then can you go about creating and growing that value for users, and for your business.
#1 Organic vs. paid users
What percentage of your new users are organic?
In a business with network effects, the share of organic users relative to paid users (the ones you spend to acquire) should increase over time. This is because as the network grows and becomes more valuable for users to join, more users should want to join on their own.
For companies with direct-side network effects (also called demand-side increasing returns), such as Facebook, the organic share will grow as people get their friends to join the platform — because their own experiences also improve as a result. For two-sided marketplaces such as Airbnb, eBay, and others, the organic share of new users grows as more suppliers (housing, sellers) and buyers want to join the network to get access — because of the potential revenue and variety of choice there.
To be clear, this doesn’t mean that paid acquisition is a bad thing; many companies including the likes of Facebook and Uber spend money to acquire new users, especially in new markets. But any company seeking to grow a sustainable business will reduce their share of paid acquisition once they reach a critical mass of users. Bottom line: as any network grows larger and its value to users grows, it should become less dependent on paid acquisition.
But there is a second layer of subtlety here, which is defining the relevant denominator of users for which the organic portion should be expanding. The right answer depends on the product and the use case: a company that has local network effects — for instance, for finding home-service providers nearby vs. globally — will show increasing organic new users, but only on a geography-by-geography basis. For small social networks designed to be used within a specific setting (like a school), the % organic should increase at the atomic unit of that particular network (i.e., the school).
#2 Sources of traffic
As the network grows, how much traffic/transactions on the network are generated internally, arising from the network itself vs. from external sources?
Just as valuable networks should organically draw more users, valuable networks should also become a destination — where users want to spend more time on the platform (or marketplace).
Measuring traffic sources is one way to help tease this apart, by separating — and tracking — how much traffic or transactions on the network is direct vs. arises from external sources. More traffic coming directly suggests users are finding the network more valuable over time as it grows.
A useful example to think about here is OpenTable, an online restaurant reservation company; initially, the typical user flow through the then-small network would be:
outside research or discovery of a restaurant > decide on a restaurant > go to that restaurant’s website > book a reservation through their OpenTable widget
As OpenTable’s network grew, it became more useful to users, since users could see the availability of every participating restaurant, vs. a specific restaurant through its own website. The share of direct traffic increased as users started their discovery process on OpenTable:
research or discover restaurant on OpenTable > decide on a restaurant > skip the restaurant’s website and book a reservation directly on OpenTable
A similar example is Medium: as the network grew and there was more content to read and more writers to follow, a greater percentage of read time originated from within the Medium site. These are also both classic examples of “come for the tool, stay for the network” type businesses.
#3 Time series of paid CAC
How much do you need to spend to acquire supply?
While paid CAC (customer acquisition cost) should theoretically decline over time in a business once the network effects “flywheel” starts accelerating, in reality, this also depends on a number of other factors: the competitiveness of marketing channels like Facebook, where prices can increase with more demand from more advertisers; the availability of substitute products; viral loops; and so on.
For example, with ridesharing, there are a variety of substitutes available to drivers — the more constrained side of the market — which has made it more expensive over time to acquire that supply. But with a company like OpenTable, which aggregated demand onto a single platform, it became cheaper to acquire restaurants over time.
By the way: it’s common to confuse network effects with virality, but the two concepts are different. Network effect businesses refer to increasing value of the product/service with each incremental user. And while network effect businesses often have a large viral growth component, which can impact CAC, the two concepts are not the same. Viral growth (users inviting other users) can exist in a non-network effect business. So, companies that are great at viral growth but don’t have network effects can grow quickly but flame out just as fast.
#4 Prevalence of multi-tenanting
How many of your users also use other similar services? How many users are active on similar services?
It’s important to understand whether your users are also using similar services, including related services where the functionality may not be exactly the same.
We’ve often observed that if a company is able to replicate a network, it can also layer on functionality that can obviate the need for another product. Even if it doesn’t wipe out the target company, such multi-tenanting can reduce usage and compress margins for all competitors. A marketplace for dog walkers and pet owners, for example, has the opportunity to move into pet health or food or other adjacent products, given it has built a network of pet owners from the core business. Facebook developed ephemeral Stories and added this feature into their various apps, including Instagram, in turn stymying the growth of Snapchat.
Measuring such multi-tenanting can be tricky — it might mean polling your users and asking whether they use another service; digging deeper into churn or declines in usage (and figuring out whether those users are moving to a different service); or simply brute-force searching for users’ profiles on other platforms! But once you see how many users are multi-tenanting, there are ways to shore up your product so users are less tempted to go somewhere else. In ride-sharing, for example (which had high multi-tenanting on both sides), companies rolled out subscriptions on the rider side and bonuses on the driver side to boost retention and reduce usage of competitors’ services.
Finally, even if you have a good sense of the overlap between your user base with another service’s, it’s important to consider how active they are: are they merely maintaining a profile, or actively using it? Having a LinkedIn profile is ubiquitous among professionals, but knowing if those users are active or not is important to know for a new startup trying to create a professional services network, so they could target areas where those users are not served well by the current product. This is a common strategy for networks to “Trojan Horse” their way into building a competing network from the underserved segment of the market first.
#5 Switching or multi-homing costs
How easy is it for users to join a new (and even a non-existent) network? How much value can users get as a new user from joining a different network?
Beyond the availability of substitutes, how easy it is for users of one network to sign up and complete the onboarding process for a competing network?
The friction involved in signing up and becoming an active user varies from product to product. Products that have an onboarding process that requires high upfront investment may find it challenging to activate prospective new users — but it also serves as a moat against competitors, because once those users are active, they’re less likely to multi-tenant. Looking at the landscape of online personal styling services, a Stitch Fix customer for instance may find it tedious to try out a different service because of the upfront investment in explaining her preferences to a new stylist; inputting information around her taste and sizing; calibrating various styles received and returned; and so on.
Conversely, if a product has a lower activation energy required of new users, it can more easily wedge its way into a market by getting users to multi-tenant and switch over: Because Uber already had millions’ of users’ credit card information for ride-sharing purposes, a user who was previously using another food delivery network could easily start using Uber Eats without much friction.
Another important consideration here is how much value can users get at the beginning when they join a new network — what’s the user experience with a cold start? For Facebook, even though users can easily join other social networks, their data, content, and networks are all on Facebook, so there’s high switching costs to inviting their network and rebuilding their social graph. On the other hand, for job listing marketplaces, an employer can easily upload their hiring specs to multiple sites and start receiving candidate applications from the get-go.
Distilling switching or multi-homing costs into a quantifiable metric can be tricky, and any metric will be quite specific to that exact business and market. Potential metrics could be the time required to complete a competitor’s onboarding flow; or the ease of getting to the minimum threshold or “magic number” for a product to be useful (e.g. 10 friends for Facebook); and so on.
#6 User retention cohorts
Is your user retention improving for newer cohorts?
The classic definition of a network effect is that the value of a product or service to a user increases with the number of other users using the same product or service. This increase in user value should therefore be reflected in user retention cohorts: newer cohorts (who experience a product when the network is larger and more useful) should have better retention for any given time period than older cohorts that joined when the network was smaller.
However, theory often differs from reality here, and we often see businesses that have declining cohort retention over time. This is because a major confounding factor to consider when evaluating user retention (metrics #6-8 on this list) is that the oldest user cohorts — especially for social network/community-based products — tend to be early adopters who are the most “ideal customers” for a product/service. Those early, often highly motivated users naturally translate into better retention cohorts for the oldest customers, rather than the newest.
Other circumstances can also change the analysis of this metric: the presence of a competitor; network effects that are hyperlocal and thus “reset” for new users in every new geography; or even negative network effects, where value to users actually decreases at a certain threshold (perhaps due to crowding or contaminants in the network).
#7 Core action retention cohorts
Is retention, as defined by users taking a core action for the product, improving for newer cohorts?
Digging deeper into the engagement funnel, you want to see if more users are taking the “core action” of your product. The core action can be one that actually corresponds to users deriving value from your product, and/or something that maps closely to your business model.
For instance, if the core action of Nextdoor is users posting content on neighborhood newsfeeds, then as the network density grows, they should expect to see improving retention as anchored on this core action. This core action retention is more telling of network effects than just measuring top-level logins or app opens.
#8 Dollar retention & paid user retention cohorts
Are newer cohorts retaining better on a dollar basis, for every given time period, than older cohorts?
Subscription and paid products need to pay attention to dollar retention and paid user retention. New user cohorts should be better retained — in terms of cohort revenue — than older cohorts. Why? Because paying for a product indicates how much users value that product, a product with network effects — which becomes more valuable over time — should have increasing dollar retention and paid user retention among newer cohorts.
For instance, as the network coverage of Angie’s List — a home services directory — improves, we’d expect to see that new user subscriber cohorts are better retained, both in terms of dollar retention as well as the number of users who remain subscribed, given the greater utility of the site.
#9 Retention by location/geography
Are participants in the oldest markets — for businesses with local network effects — better retained, than those in newer markets?
For local network effect businesses, the network effects exist on a per-market basis, and “resets” for new geographies. For Care.com users in Charlotte, for example, the presence of more babysitters available in New York City doesn’t impact the user experience; but having more babysitters available locally does improve the usefulness of the network there.
As each geography matures and builds network density, retention should improve in those markets. Thus, the oldest or most established markets tend to have better retention than newer markets. We see this in practice in data shared by almost every local network effect business.
#10 Power user curves (aka L7 & L30 charts)
Are users shifting to the right side of the power user curve? In other words, are they becoming more engaged over time?
Power users drive some of the most successful companies, by contributing a ton of value to the network. While DAU/MAU — dividing daily active users by monthly active users — is a common metric for measuring engagement, it has its shortcomings, and power user curves provide a more nuanced way to understand user engagement.
In short, power user curves (commonly called L30 charts for 30 days of use, or L7 charts for 7 days of use) are histograms of users’ engagement, showing the total number of days users were active in doing a particular action in a given timeframe. In analyzing network effect businesses, seeing how often users take a specific action on a cohort basis allows you to see whether a product is really gaining utility with more users — aka the network effect. If a product is indeed more valuable with more users, then that should be reflected in a growing share of users shifting to higher-frequency engagement buckets, or a more right leaning power user curve, over time.
#11 Match rate (aka utilization rate, success rate, etc.)
How successfully can the two sides of the marketplace find each other?
The job of any marketplace is to facilitate the matching of supply with demand. It’s therefore important to measure your successful “match rate” — the rate at which buyers can find sellers, and vice versa. How to define this metric depends on the unique business.
Match rates examples for particular businesses include:
- Driver utilization time for ridesharing — what % of the time are drivers driving around with a passenger, vs. empty?
- How often are employers actually filling their posted role in job marketplaces? And how often are job seekers finding jobs?
A related metric is to measure “zeros”, or unsuccessful matches. For ridesharing, what percentage of users open the app but don’t end up requesting a ride? Those “zeros” could be due to too long of a wait time, surge pricing, or something else — all instances the marketplace was unable to clear demand. Marketplace operators should identify reasons why matches don’t happen and take steps to remove or reduce these blockers through growing and incentivizing the more constrained side of the marketplace, improving product design, and other mechanisms.
This metric is also closely related to the concept of multi-tenanting described above. If match rate is low, then users will naturally be incentivized to go elsewhere and use other products. For instance, it’s common for employers to post their job listings on a variety of sites — their own website, LinkedIn, Indeed, as well as other networks — simply because no single network has a high enough match rate. If there’s even incremental revenue potential or even just minimum utility, multi-tenanting will take place; just think about all of the delivery marketplace stickers you see in any given restaurant’s window!
#12 Market depth
Is there enough supply and does it fit users’ needs?
The concept of “offer depth” or market depth originated from financial markets, where it’s defined as the market’s ability to sustain relatively large orders without price movements. The higher the number of buy and sell orders at each price, the greater the depth of the market.
For consumer marketplaces, it’s important to measure market depth because it directly impacts the user experience. For heterogeneous supply marketplaces (where each supplier is different), market depth determines whether users will be able to find a match. When users open products like OfferUp or Airbnb, how many listings will they see, and how likely will they be to find an item they want to buy or home they want to rent? For homogenous supply marketplaces, market depth impacts ease of use. When users open Lime, how many bikes/scooters will they see near them? The greater the market depth, the easier (and less user effort required, in terms of walking) it is to use Lime.
One of the primary jobs of any marketplace business is to reduce search costs — making it easy for participants to find and match with the other side. Failing to do this can result in a marketplace with negative network effects, where too much supply actually causes challenges in discovery. As consumers, we experience this as decision fatigue, or a paradox of choice. Conversion rates could fall in this scenario.
A note on heterogeneous vs. homogeneous supply: “homogeneous supply” marketplaces typically hit an asymptote in network effects, where the value to users eventually plateaus with greater market depth. For instance, if there were 6 Lime scooters on a city block near me, this is no more valuable than if there were only 4 or 5 scooters available for me to use in my vicinity — user value is unchanged despite the addition of more supply. On the other hand, for heterogeneous marketplaces, there is no asymptote because every node on the supply side is different and potentially can add greater value. In the Airbnb example, a user’s tastes may be quite specific, so every additional listing on the platform is useful to see.
#13 Time to find a match (or inventory turnover, or days to turn)
How long does it take for supply and demand to match?
Typically, marketplaces have a curve for match rate, where over a long time horizon, a greater share of inventory clears. For product marketplaces, this is commonly referred to as inventory turnover.
The inverse is days to turn, and this metric is more applicable for more traditional marketplaces, where the matching happens via users opting in — one side creates a listing and the other responds — in contrast to on-demand marketplaces, which do matching in a centralized, algorithmic (and less visible to users) way.
For instance, for job marketplaces, how long does it take an employer to find an employee? How long does it take to receive the first application? For P2P marketplaces, how long does it take for each side to engage in a transaction? For Thumbtack, how long does it take users to receive the first quote? How long does it take on OfferUp for a seller to sell their product?
#14 Concentration or fragmentation of supply and demand
How concentrated is the marketplace on the supply and demand sides?
Marketplaces where there is greater fragmentation on the supply and demand sides are more valuable and defensible. This means no participants on the demand or supply sides disproportionately account for a high share of transactions, which makes the business more sustainable and diversified. If demand or supply is too concentrated on a marketplace, there’s risk that a large buyer or seller can take a large share of transactions with them if they decide to leave the platform.
There’s also greater value when a marketplace aggregates fragmented goods or providers, as those would otherwise have been more difficult to discover and access. This is basically like taking the advantages of a long tail (more variety and niches) and making it as easy to find as the head of the tail (beyond just popular hits).
Marketplaces can gauge concentration by measuring the % of GMV the top X sellers or buyers account for (e.g. the share of GMV each grocery chain contributes, in the case of Instacart).
#15 Pricing power
How much are you able to charge for your product? What would your customers be willing to pay to stay on the network?
As participants receive greater value from the network, they are willing to pay more to have access to network, in the form of subscriptions, listing fees, take rates, or other monetization mechanisms.
Over the lifetime of a network effects business, the business can evolve from not being monetized at all, or potentially even subsidizing demand or supply; to turning on monetization; to having the ability to increase prices with minimal churn on either side.
#16 Unit economics
How is the business doing, basically?
Improved network effects often appear in improved unit economics over time. This is a result of declining incentives that businesses need to offer to different sides of the market, lower share of paid users, and overall improvement in pricing power.
For businesses with local network effects, the impact of network effects should show up in unit economics over time, on a market-by-market basis. This is because in a given market, CAC should decrease and the organic share of users should grow over time. For businesses like Thumbtack or Instacart, which have network effects at the local level, tracking the unit economics over time per market is helpful because you’ll see the relationship between market age, network density, and profitability.
The most successful companies and products of the internet era have all been predicated on the concept of network effects, where the network becomes more valuable to users as more people use it. This is as true of companies like Amazon and Google as it is for open source projects like Wikipedia and some cryptocurrencies. At its core, the theory behind network effects suggests that platforms and products with network effects get better as they get bigger — not just in value to users, but also in accruing more resources to improve their product, thus strengthening the “flywheel”.
But recently, reality seems to be diverging from theory.
Instead of seeing winner-take-all markets, we’re seeing all kinds of network effects companies — from messaging apps to sneaker marketplaces — splitting markets among multiple players. Furthermore, even companies that appear to have initially won the market and seem to have established a deep moat — from dating apps to trading platforms — are struggling to maintain their position against copycats and new entrants. Just look at what Instagram Stories is doing to Snap, surpassing it recently even among its dedicated teen demographic.
Does all this mean that network effects as we know them are dead? No, but they’re more dynamic than ever.
While we know that not all network effects are created equal, they don’t evolve equally either. Every product has different types of network effects that mature and develop differently over time. If anything, most network effects businesses are changing faster than ever before. So how can entrepreneurs and founders navigate this era of seemingly diminishing network effects? The trick is to know what your network effects look like today, but also project how they’ll evolve over time. To that end, you’ll need to understand three aspects of your company and how they could change going forward: 1) your value proposition, 2) your users/inventory, and 3) your competitive ecosystem. Otherwise you could get caught flat-footed, claiming that network effects are dead.
Here’s some principles for forecasting future network effects beyond a present-day, static view. Then, once you know you have network effects, see this post for how to measure and keep them.
1) Value prop: Not all products are created equal
A company’s or product’s network effects don’t always remain on an increasing-returns (or even straight) line trajectory as it grows; it could asymptote, or hit an inflection point and even reverse. The key for founders is to know what value proposition drives their network effects, understand whether they’re weak or strong, and then pay close attention to how they will evolve — especially as you iterate your way to new value propositions and additional layers of product-market fit.
Let’s take a look at a few examples:
Ridesharing. Within any geography, driver supply and passenger demand reinforce each other, so more drivers mean lower wait times, which means more passenger demand, which means more drivers want to drive… the flywheel spins! But there’s a hitch to that value proposition: Once you hit the 5-minute wait time in a particular geography, riders are indifferent to whether there are more drivers available in the network. And once multiple platforms hit that level of relative liquidity (e.g., enough drivers to satisfy the riders 5-minute wait time), the specific platform matters less to riders. In those markets, they’ll need to compete on other vectors, like brand/reputation, price, user experience, loyalty programs, and more. In other words, the value proposition needs to evolve.
Social lending. Sometimes changes in network effects are driven not by relative liquidity (like wait times), but by absolute liquidity (like the number of people in a network). Take the case study of Frank, which let people borrow money from and lend money to friends and family [full disclosure: one of us co-founded Frank]. Early on, more friends in Frank groups meant more demand and more liquidity, which created a bigger incentive for people to join those groups. But once a group had more than 7 people, they became less likely to lend or borrow: turns out people only have ~7 friends/family members they have that level of trust with! The network effects in this case went from positive to negative as an individual’s network outgrew the value proposition. This pattern has also held true for a number of other highly social products.
Social networks. As Facebook’s use case went from sharing status updates with friends to surfacing news and content, the network effects weakened. Too many friends/followers meant people didn’t feel as comfortable sharing personal content, and the experience shifted more to news and public content. The value proposition and network effects shifted accordingly — from social networking to social media. Importantly, adding additional nodes in the context of media discovery was less valuable than it was in the earlier days of a pure social network. So the shift in value proposition meant the network effect curve hit an inflection point. And while some features (e.g. curation algorithms, timelines) helped defer this tipping point by letting users add more friends while still letting the content feel relevant, eventually the content mix, the value proposition, and the network effects all shifted.
Decentralized platforms. If one thinks of bitcoin, for instance, as digital gold, then the network effect would be that more buyers/sellers mean more liquidity which increases the value of the platform for all. But if one thinks of bitcoin as a payments platform, then more is not necessarily better as long as it experiences network congestion or other friction. It’s an interesting case to consider (not for investment purposes), simply because it’s an example of how different value propositions for the same platform can strengthen or weaken network effects accordingly. It’s also a great example of how additional features (e.g., scaling, increased throughput, and improved transaction speeds) can help the value proposition evolve, shift, or even create a new trajectory of network effects.
The goal of sharing the above examples is to show both the nuances and evolution of network effects. If you’re not paying attention to these factors, you could be left believing that network effects just don’t exist anymore in a particular business, when it might be a matter of unlocking new value.
2) Users and Inventory: Not all users are created equal
The type of users and inventory your product or platform has today, and the types you’re adding, are fundamental in understanding and projecting the trajectory of your network effects.
Commoditized vs. differentiated supply
An important factor in projecting network effects — especially in two-sided platforms/marketplaces — is whether users/inventory are commoditized or differentiated.
In ridesharing, the customer (rider) is relatively agnostic to the underlying service provider/inventory because they perceive the supply (drivers/cars/transport) as interchangeable and therefore commoditized. Platforms with relatively commoditized inventory — from on-demand storage companies to delivery companies — are more likely to see their network effects asymptote once they reach a base level of liquidity. For a category like ridesharing, moving into adjacent businesses (like Lyft has done with their healthcare initiative or Uber has done with food delivery) allows differentiated — yet still substitutable — inventory, potentially increasing the strength of the network effect.
Platforms/marketplaces with more differentiated inventory have stronger and longer-lasting network effects, because they have a diversity of inventory that suits the unique preferences of customers (while maintaining just-enough substitutability across that inventory as well). For example, AirBnB can show users every iteration of lodging from $225-$325/night in Los Angeles, which overlaps with someone else’s search for something that costs $150-$250 and has a both a balcony and a hot tub. The platform is therefore more valuable on both sides of the marketplace than a site that just shows a commoditized set of standard and executive rooms. The network effects remain strong not only because it reaches a base level of liquidity across all these different types of inventory (making them valuable to more users), but because it also continues to see increasing returns with new supply.
But the more differentiated the inventory, the more the platform needs to do a good job of curation and matching. That in and of itself also increases the overall defensibility of the platform, and keeps the network effects curve strong over time.
Type of incremental user
Beyond the commoditized nature of users and inventory, however, not all members of a given network are equal. Some are more — or less — valuable than others. For instance, a restaurant that is very popular and located nearby a lot of users adds more value to the OpenTable network than a restaurant with bad food in the middle of nowhere.
When you forecast out your network effects — and more importantly, your growthstrategy for acquiring and engaging more users — you will need to pay attention to the incremental users you’re likely to attract. Are they network “contaminants”, “neutrals”, or “contributors”? For a social network, adding a troll that disengages other users is a pollutant who removes value. Adding a lurker is neutral since that person doesn’t add or subtract any value from the network. Adding a great content producer contributes an enormous amount of value to the network.
So, making sure to incent the users you want while disincenting the ones you don’t want, is key. This is why most great platforms also invest heavily in curation mechanisms to screen and remove bad inventory/users (e.g., Wikipedia’s editors, Airbnb’s reviews/onboarding, etc.). Unfortunately, these screening mechanisms don’t always work and sometimes the cost of finding strong contributors becomes very high, so the calculus of growth relative to cost matters a lot here.
3) The Competition: Not all markets are created equal
The nature of your market as well as competitors and substitutes is also critically important to understanding and forecasting network effects.
While network effects businesses tend to be more defensible at scale, they are not immune to competition. But for these types of businesses it’s not just a matter of figuring out who your direct competitors are — you also need to think about the network overlap. If someone else has a similar network to yours, there’s always existential risk they’ll move into your market. Because they have a similar network already, they’ll more easily be able to enter your space (Instagram’s foray into Snapchat-like disposable “Stories” is a good example of this). This is also true where the competition may already have registered a superset of your network (e.g., DoorDash and Uber Eats; Didi and Uber in China).
Low switching costs to competitors can also lower network effects. Seamless sign-up and onboarding is usually great for adding users to your product, but if your competitors have that same seamless onboarding, users might multi-tenant. It’s easy to use multiple dating apps or maps products because of low barriers to entry and switching costs.
Multi-tenanting to meet demand
Network effects are weakened when users are unable to use a single platform to accomplish their goals. Jobs marketplaces are a good example here: companies are likely to list their job openings on multiple hiring platforms (i.e., multi-tenant), since hiring is a critical part of running a business that absolutely needs to be fully fulfilled, and no single platform is likely to fulfill all of their hiring needs. Where one side of a platform is multi-tenanting, there will usually be increased pressure on the operator — in terms of pricing, features, and necessary liquidity — which can turn the economics upside down.
We go into detail on how to measure all of these in our other post, 16 metrics for measuring network effects: https://a16z.com/2018/12/13/16-metrics-network-effects/.
* * *
Fast followers can move faster than ever. Instagram Stories can challenge Snap in, well, a snap. Hiring marketplaces can get thousands of employers in no time. The API-ization of everything makes it easier to do anything and everything: Stripe alone means that every marketplace can incorporate payments in an hour, whereas previously, eBay had some protection by building and acquiring its own payment systems.
The increasing speed of product iteration, the pace at which networks can scale, and the ease with which competitors can get started has therefore dramatically changed how we project network effects in businesses. Instead of winner-take-all markets where early movers may have once had a seemingly lasting advantage, network effects change more quickly than ever. Especially where specific factors — an asymptotic value proposition, network overlap, increasing number of contaminants, etc. — can lower the platform’s ability to generate a sustainable network effect in the future.
This is not to depress anyone! Network effects will continue to underpin the most impactful software businesses. It’s merely a reminder to founders and other product builders to be aware of what will change and why, so you can do more to plan and address these issues instead of being blindsided by them. Network effects aren’t dead, but they’re more dynamic than ever. By understanding what your network effects look like today and where they’re going tomorrow, founders can design network effects and other moats with intentionality, instead of being tossed on the winds of change.
Long live your network effects!
[I co-wrote this essay with my a16z partner Andrew Chen. We’re excited about the future of marketplaces and the service economy. Hope you enjoy this glimpse into our thinking, and let us know if you have any feedback!]
Goods versus Services: The next trillion dollar opportunity
Marketplace startups have done incredibly well over the first few decades of the internet, reinventing the way we shop for goods, but less so for services. In this essay, we argue that a breakthrough is on its way: The first phase of the internet has been about creating marketplaces for goods, but the next phase will be about reinventing the service economy. Startups will build on the lessons and tactics to crack the toughest and largest service industries — including regulated markets that have withstood digital transformation for decades. In doing this, the lives of 125 million Americans who work in the services-providing industries will join the digital transformation of the economy .
In the past twenty years, we’ve transformed the way people buy goods online, and in the process created Amazon, eBay, JD.com, Alibaba, and other e-commerce giants, accounting for trillions of dollars in market capitalization. The next era will do the same to the $9.7 trillion U.S. consumer service economy , through discontinuous innovations in AI and automation, new marketplace paradigms, and overcoming regulatory capture.
Amazon and eBay were among the first major e-commerce websites that made it possible for consumers to buy things without having to step into a physical store. Since then, the rise of e-commerce platforms like Shopify, discovery platforms like Instagram, and companies that provide fulfillment and logistics services have lowered the barriers to entry to starting an online store, leading to an explosion in the goods available online.
In contrast, the service economy lags behind: while services make up 69% of national consumer spending, the Bureau of Economic Analysis estimated that just 7% of services were primarily digital, meaning they utilized internet to conduct transactions .
We propose that a new age of service marketplaces will emerge, driven by unlocking more complex services, including services that are regulated. In this essay, we’ll talk about:
- Why services are still primarily offline
- The history of service marketplace paradigms
- The Listings Era
- The Unbundled Craigslist Era
- The “Uber for X” Era
- The Managed Marketplace Era
- The future of service marketplaces
- Regulated services
- Five strategies for unlocking supply in regulated markets
- Future opportunities
Let’s start by looking at where the service economy is right now and why it’s resisted a full scale transformation by software.
Software is eating the service economy, but it’s been slow
We’ve all had the experience of asking friends for recommendations for a great service provider, whether it be a great childcare provider, doctor, or hair stylist. Why is that? Why aren’t we discovering and consuming these services in the same digital way we’ve come to expect for goods?
Despite the rise of services in the overall economy, there are a few reasons why services have lagged behind goods in terms of coming online:
- Services are complex and diverse, making it challenging to capture relevant information in an online marketplace
- Success and quality in services is subjective
- Fragmentation — small service providers lack the tools or time to come online
- Real-world interaction is at the heart of services delivery, which makes it hard to disaggregate parts of a purchase that might be done online
Let’s unpack each reason below:
First, on the complexity and diversity of services, services are performed by providers who vary widely, unlike goods which are manufactured to a certain spec. Even the names of services can vary: what one home cleaning service calls a “deep clean” can be different from another provider’s definition. This lack of standardization makes it difficult for a service marketplace to capture and organize the relevant information.
Second, services are often complex interactions without a clear yardstick of success or quality. The customer experience of a service is often subjective, making traditional marketplace features like reviews, recommendations, and personalization more difficult to implement. Sometimes just getting the job completed (as in rideshare) is sufficient to earn a 5-star review, whereas other higher-stakes services, like childcare, have complex customer value functions, including safety, friendliness, communicativeness, rapport with child, and other subjective measures of success.
Third, small service providers often lack the tools or time to come online. In many service industries, providers are small business owners with low margins; contrast this with goods manufacturing where there are economies of scale in production, and thus consolidation into large consumer products companies. As a result of industry fragmentation, service providers often don’t have time or budget to devote to key business functions, such as responding to customer requests, promoting and marketing themselves, maintaining a website, and other core functions. While major e-commerce platforms have taken on the role of distribution, merchandising, and fulfilling orders for goods, there are few platforms that service providers can plug into to manage their businesses and reach customers.
Fourth, real-world interaction is central to services, which can pull other steps of the services funnel into the offline world as well. Many services are produced and consumed simultaneously in real-world interactions, whereas goods entail independent stages of production, distribution, and consumption. The various stages of the goods value chain can be easily unbundled, with e-commerce marketplaces comprising the discovery, transaction, and fulfillment steps. Conversely, since the production and consumption of services usually occur simultaneously offline, the discovery, distribution, and transaction pieces are also often integrated into the offline experience. For instance, since getting a haircut entails going to a salon and having interactions with the providers there, the stages of the value chain that precede and follow that interaction (discovery, booking, and payment) also often get incorporated into the in-person experience.
The 4 Eras of Service Marketplaces, and What’s Next
There have been 4 major generations of service marketplaces, but coverage of services and providers remains spotty, and many don’t provide end-to-end, seamless consumer experiences. Let’s zoom out and talk through each historical marketplace paradigm, and what we’ve learned so far.
1. The Listings Era (1990s)
The first iteration of bringing services online involved unmanaged horizontal marketplaces, essentially listing platforms that helped demand search for supply and vice versa. These marketplaces were the digital version of the Yellow Pages, enabling visibility into which service providers existed, but placing the onus on the user to assess providers, contact them, arrange times to meet, and transact. The dynamic here is “caveat emptor” — users assume the responsibility of vetting their counterparties and establishing trust, and there’s little in the way of platform standards, protections, or guarantees.
Craigslist’s Services category is the archetypal unmanaged service marketplace. It includes a jumble of house remodeling, painting, carpet cleaners, wedding photographers, and other services. But limited tech functionality means that it feels disorganized and hard to navigate, and there’s no way to transact or contact the provider without moving off the platform.
2. The Unbundled Craigslist Era (2000s)
Companies iterated on the horizontal marketplace model by focusing on a specific sub-vertical, enabling them to offer features tailored to a specific industry.
Angie’s List, a home services site founded in 2005, carves off Craigslist’s household services category. The platform has features including reviews, profiles, certified providers, and an online quote submission process. But the marketplace doesn’t encompass the entire end-to-end experience: users turn to Angie’s List for discovery, but still need to message or call providers and coordinate offline.
Like previous listing sites, these platforms in this era try to use the “wisdom of the crowds” to promote trust. These platforms have a network effect in that more reviews means more users and more reviews. But user reviews have their limitations, as every user has a unique value function that they’re judging a service against. Without standardized moderation or curation, and without machine learning to automate this process, customers have the onus of sifting through countless reviews and selecting among thousands of providers.
3. The ‘Uber for X’ Era (2009-2015)
In the early 2010s, a wave of on-demand marketplaces for simple services arose, including transportation, food delivery, and valet parking. These marketplaces were enabled by widespread mobile adoption, making it possible to book a service or accept a job with the tap of a button.
Companies like Handy, Lugg, Lyft, Rinse, Uber and many others made it efficient to connect to service providers in real-time. They created a full-stack experience around a particular service, optimizing for liquidity in one category. For these transactions, quality and success were more or less binary — either the service was fulfilled or it wasn’t — making them conducive to an on-demand model.
These platforms took on various functions to establish an end-to-end, seamless user experience: automatically matching supply and demand, setting prices, handling transactions, and establishing trust through guarantees and protections. They also often commoditized the underlying service provider (for instance, widespread variance on the driver side of rideshare marketplaces is distilled into Uber X, Uber Pool, Uber Black, Uber XL, etc.).
Unlike the previous generations of marketplaces, in which the provider ultimately owns the end customer relationship, these on-demand marketplaces became thought of as the service provider, e.g. “I ordered food from DoorDash” or “Let’s Uber there,” rather than the underlying person or business that actually rendered the service.
Over time, many startups in this category failed, and the ones that survived did so by focusing on and nailing a frequent use case, offering compelling value propositions to demand and supply (potentially removing the on-demand component, which wasn’t valuable for some services), and putting in place incentives and structures to promote liquidity, trust, safety, and reliability.
4. The Managed Marketplace Era (Mid-2010s)
In the last few years, we’ve seen a rise in the number of full-stack or managed marketplaces, or marketplaces that take on additional operational value-add in terms of intermediating the service delivery. While “Uber for X” models were well-suited to simple services, managed marketplaces evolved to better tackle services that were more complex, higher priced, and that required greater trust.
Managed marketplaces take on additional work of actually influencing or managing the service experience, and in doing so, create a step-function improvement in the customer experience. Rather than just enabling customers to discover and build trust with the end provider, these marketplaces take on the work of actually creating trust.
In the a16z portfolio, Honor is building a managed marketplace for in-home care, and interviews and screens every care professional before they are onboarded and provides new customers with a Care Advisor to design a personalized care plan. Opendoor is a managed marketplace that creates a radically different experience for buying and selling a home. When a customer wants to sell their home, Opendoor actually buys the home, performs maintenance, markets the home, and finds the next buyer. Contrast this with the traditional experience of selling a home, where there is the hassle of repairs, listing, showings, and potentially months of uncertainty.
To compensate for heavier operational costs, it’s common for managed marketplaces to actually dictate pricing for services and charge a higher take rate than less-managed marketplace models.
Managed marketplaces are a tactic to solve a broader problem around accessing high-quality supply, especially for services that require greater trust and/or entail high transaction value. If we zoom out further, there are many more categories of services that can benefit from managed models and other tactics to unlock supply.
What’s Next: The Future of Service Marketplaces (2018-?)
We think the next era of service marketplaces have potential to unlock a huge swath of the 125 million service jobs in the US. These marketplaces will tackle the opportunities that have eluded previous eras of service marketplaces, and will bring the most difficult services categories online — in particular, services that are regulated. Regulated services — in which suppliers are licensed by a government agency or certified by a professional or industry organization — include engineering, accounting, teaching, law, and other professions that impact many people’s lives directly to a large degree. In 2015, 26% of employed people had a certification or license .
Regulation of services was critical pre-internet, since it served to signify a certain level of skill or knowledge required to perform a job. But digital platforms mitigate the need for licensing by exposing relevant information about providers and by establishing trust through reviews, managed models, guarantees, platform requirements, and other mechanisms. For instance, most of us were taught since childhood never to get into cars with strangers; with Lyft and Uber, consumers are comfortable doing exactly that, millions of times per day, as a direct result of the trust those platforms have built.
Licensing of service professions creates an important standard, but also severely constrains supply. The time and money associated with getting licensed or certified can lock out otherwise qualified suppliers (for instance, some states require a license to braid hair or to be a florist), and often translate into higher fees, long waitlists, and difficulty accessing the service. The criteria involved in getting licensed also do not always map to what consumers actually value, and can hinder the discovery and access of otherwise suitable supply.
Five Strategies to Unlock Industries
We’re starting to see a number of startups tackling regulated services industries. As with each wave of previous service marketplaces, these new approaches bring more value-add to unlock the market, with variations in models that are well-suited to different categories.
The major approaches in unlocking supply in these regulated industries include:
- Making discovery of licensed providers easier
- Hiring and managing existing providers to maintain quality
- Expanding or augmenting the licensed supply pool
- Utilizing unlicensed supply
- Automation and AI
1) Making discovery of licensed providers easier
Some startups are tackling verticals that lack good discovery of licensed providers. Examples include Houzz, which enables users to search for and contact licensed home improvement professionals, and StyleSeat, which helps users find and book beauty appointments with licensed cosmetologists.
2) Hiring and managing existing providers to maintain quality
Companies can raise the quality of service by hiring and managing providers themselves, and by managing the end-to-end customer experience. Examples are Honor and Trusted, managed marketplaces for elder care and childcare, respectively, which employ caregivers as W-2 employees and provide them with training and tools. In the real estate world, Redfin agents are employees whose compensation is tied to customer satisfaction, unlike most real estate agents who are independent contractors working on commission.
3) Expanding or augmenting the licensed supply pool
Expanding the licensed supply pool can take the form of leveraging geographic arbitrage to access supply that’s not located near demand. Decorist, Havenly, Laurel & Wolf, and other online interior design companies enable interior designers around the world to provide design services to consumers without physically visiting their homes (yes, in many parts of the US interior design requires a license!). With improvements in real-time video, richer telepresence technologies, and better visualization technologies, more synchronous services are also shifting from being delivered in-person to online. Outschool and Lambda School are examples of de-localizing instruction, enabling teachers and students to participate remotely while preserving real-time interaction.
Another approach is to help suppliers navigate the certification process. a16z portfolio company Wonderschool makes it easier for individuals to get licensed and operate in-home daycares.
Lastly, there’s the approach of augmenting certified providers so they can serve more customers. Fuzzy, an in-home veterinary service, uses AI and vet technicians to augment the productivity of licensed veterinarians; and a16z portfolio company Atrium builds automation and workflow management to provide efficiency gains in the legal industry.
4) Utilizing unlicensed supply
Some companies utilize unlicensed supply — notably Lyft, Uber, and other peer-to-peer rideshare networks. Another example is Basis, a managed marketplace for guided conversations with trained but unlicensed specialists to help people with anxiety, depression and other mild to moderate mental health issues.
In the pet space, Good Dog is a marketplace that brings together responsible pet breeders and consumers looking for a dog. Going beyond existing breeder licensing, which the company felt didn’t map to what consumers valued, Good Dog established its own higher set of standards and screening process in conjunction with veterinary and academic experts.
5) Automation and AI
Other startups automate away the need for a licensed service provider altogether. These include MDacne, which uses computer vision to diagnose and treat acne; and Ike Robotics and other autonomous trucking startups which remove the need for a licensed truck driver.
Opportunities for companies addressing regulated services
The last twenty years saw the explosion of a number of services coming online, from transportation to food delivery to home services, as well as an evolution of marketplace models from listings to full-stack, managed marketplaces. The next twenty years will be about the harder opportunities that software hasn’t yet infiltrated — those filled with technological, operational, and regulatory hurdles — where there is room to have massive impact on the quality and convenience of consumers’ everyday lives.
The services sector represents two-thirds of US consumer spending  and employs 80% of the workforce . The companies that reinvent various service categories can improve both consumers’ and professionals’ lives — by creating more jobs and income, providing more flexible work arrangements, and improving consumer access and lowering cost.
The companies mentioned in this essay just scratch the surface of regulated industries. You can imagine a marketplace for every service that is regulated, with unique features and attributes designed to optimize for the customer and provider needs for that industry. (A full list of regulated professions in the US can be found here.) We fully expect more Airbnb- and rideshare-sized outcomes in the service economy.
If you’re a founder who is looking to take on the challenge of tackling more complex services and bringing them online, we’d love to hear from you. Thank you for reading!
 Bureau of Labor Statistics, Employment by major industry sector
 Bureau of Economic Analysis, Personal Consumption Expenditures by Major Type of Product
 Bureau of Economic Analysis, Digital Economy
 Bureau of Labor Statistics, Workers in healthcare occupations most likely to have certifications or licenses in 2015
 Bureau of Economic Analysis, Personal Consumption Expenditures by Major Type of Product
 Bureau of Labor Statistics, Employment by Major Industry Sector
/ Thread: There are over 400 startups trying to be the next Warby Parker, but history shows that 90%+ of e-commerce companies will fail. What separates the successes from the failures? Here are 5 things I look for to figure out if an e-commerce startup is a good opportunity –>
/1. Does it have defensible, scalable acquisition channels?
2. Is it operating in a category that is well-suited to brand building, and if so, has it built a brand that people love and trust?
3. Is it selling a unique product that not everyone can offer?
/4. Does the business have network effects?
5. Are there economies of scale that can be captured ahead of later entrants?
/ Unpacking each one of these:
1. Customer acquisition: It’s unsustainable to rely heavily on paid acquisition channels to grow, as the margin ends up being bid away.
/ Instead, does the company have unique, scalable ways to reach new consumers, e.g. a devoted community who spreads the word, or exclusive distribution channels?
/ Dollar Shave Club’s launch video was a major accelerant for customer growth, and cost just $4500 to produce. While viral videos aren’t an uncrossable moat, it was an advantage that helped build up brand awareness cheaply, and enabled DSC to reach scale first.
/2. Brand: Brand as defensibility exists for a lot of companies–but not all product categories are suited to building strong brands, and brand defensibility is also difficult to assess in the time frames that map to venture investing.
/Brand as defensibility works better in categories where there is a big emotional component to the purchase, where a sense of identity or community is intertwined with the brand, e.g. categories like health, cosmetics, or anything aspirational.
/ Whitelabel a Glossier/Chanel/La Mer product, and chances are women won’t covet them nearly as much, even if the contents inside are the same. See my tweetstorm on the shifting power balance between brands & aggregators–and where the opportunities are:
/3. Product: Does the startup have defensibility in terms of creating a proprietary product that others don’t have the ability to create, whether that’s derived from better design, deeper customer understanding, or manufacturing moats?
/ As an example, Hubble Contacts found that the vast majority of the contact lens market is controlled by a few manufacturers. Hubble was able to establish exclusive supplier relationships that blocked other contact lens startups from scaling in the US.
/4. Network effects: There are various flavors of network effects in e-commerce. Marketplaces, obviously, improve with more buyers and sellers.
/ But even a single retailer or brand can have network effects. Stitch Fix’s algorithms for predictions and recommendations improve each time a customer reviews the items their stylist chose for them, leading to greater retention and LTV.
/5. Scale effects: Amazon is the archetypal example of a retailer that thrives on economies of scale. As a startup example, Rent the Runway purchases expensive designer apparel and accessories, amortizing that cost over multiple rentals from different customers.
/With greater scale, RTR can increase utilization, expand its inventory, and lower prices, making it harder for another women’s rentals startup to compete.
/Not every company will be a ‘yes’ to all of the above, but more of these being true can indicate a stronger opportunity.
/There’s also interactions between these: Strength in one regard can compensate for the absence of another. A manufacturing moat in a large category means establishing a beloved brand is less important. A strong distribution advantage can be investible in and of itself.
/Obviously, building a great company and making investment decisions is more nuanced than just 5 factors. It’s apparent that there’s still tons of exciting opportunities left to build great e-commerce companies, and I’d love to chat if you’re working on one!
The importance of power users
Power users drive some of the most successful companies — people who love their product, are highly engaged, and contribute a ton of value to the network. In ecommerce marketplaces it’s power sellers, in ridesharing platforms it’s power riders, and in social networks it’s influencers.
All companies want more power users, but you need to measure them before you can find (and retain) them. While DAU/MAU — dividing daily active users (DAUs) by monthly active users (MAUs or monthly actives) — is a common metric for measuring engagement, it has its shortcomings.
Since companies need a richer and more nuanced way to understand user engagement, we’re going to introduce what we’ll call the “Power User Curve” — also commonly called the activity histogram or the “L30” (coined by the Facebook growth team). It’s a histogram of users’ engagement by the total number of days they were active in a month, from 1 day out of the month to all 30 (or 28, or 31) days. While typically reflecting top-level activity like app opens or logins, it can be customized for whatever action you decide is important to measure for your product.
The Power User Curve has a number of advantages over DAU/MAU:
- It shows if you have a hardcore, engaged segment that’s coming back every day.
- It shows the variability among your users: some are slightly engaged, whereas others are power users. Contrast this with DAU/MAU: it’s a single number and so blurs this variance.
- When mapped to cohorts, Power User Curves let you see if your engagement is getting better over time — which in turn helps assess product launches and performance of other feature changes.
- Power User Curves can be shown for different user actions, not just app opens. This matters if the core activity that matters for your product is deeper in the funnel.
In other words, while the DAU/MAU gives you a single number, the Power User Curve gives entrepreneurs several avenues of analysis to assess their product’s engagement to the most addicted users — in a single snapshot, over time, and also in relation to monetization. This is useful. So how does it work?
The Power User Curve will “smile” when things are good
The shape of the Power User Curve can be left-leaning or smile-like, all of which means different things. Here’s a smile:
The Power User Curve above is for a social product, and shows the characteristic smile shape that indicates there’s a group of highly engaged users using the app daily or nearly daily. Social products with frequent user engagement like this lend themselves well to monetization via ads—there’s enough users returning frequently that the impressions can support an ad business. Remember that Facebook would have a very right-leaning smile, with 60%+ of its MAUs coming back daily.
What matters is that, over time, the platform is able to retain and grow its power users: successive Power User Curves should ideally show users shifting over more to the right side of the smile. As the density of the network grow, and with stronger network effects, it’s expected that there’s more reason for users to return on a daily basis.
The Power User Curve can show when strong monetization is needed
Let’s look a different example, which doesn’t smile:
This Power User Curve of a professional networking product looks quite different than that of a social product. It’s left-weighted with a mode of just 1 day of activity per month, and decays rapidly after those few days. There’s no power users. But this light engagement can be okay — not every company needs to have a smile-shaped Power User Curve, just as not every product category necessarily lends itself to an ultra-high DAU/MAU.
When there’s low engagement, what matters is that the company has a way to extract enough value from users when they areengaged. Think about an investing product like Wealthfront or networks like LinkedIn — few users are likely to actively check it on a daily basis, but that’s ok, since they have business models that aren’t tied to daily usage.
CEOs of such companies should therefore,think about: Is there a way to create revenue streams where the business can still monetize effectively despite users’ infrequent engagement? Or, who are the users using this product more frequently, and how can I get more of them? Is there something about the product — e.g. onboarding, the core experience, etc. — where a significant chunk of the user base isn’t experiencing the ‘aha moment’ that makes them “get” the product, and therefore not getting value from it right now (and if so how to get there)?
Some products should be analyzed in a 7 day timeframe – like SaaS/productivity – and others on 30 days
Another flavor of the Power User Curve is a histogram of users’ engagement for a 7-day period, also commonly called L7. The 7 day Power User Curve shows weekly actives, not monthly actives. Plotting this version can make sense if your product naturally follows a weekly cycle, for instance, if it’s a productivity/work-related product that users engage with Monday through Friday. B2B SaaS products will often find it useful to show this version, as they want to drive usage during the work week.
Note that using DAU/MAU wouldn’t be the appropriate metric for this product as it’s not designed to be a daily use product. You can also see there’s actually a smile curve through 5 days, but fewer users are using it 6-7 days, which makes sense for the power users of a workweek product like this.
CEOs of such product companies should therefore want to understand: Who are the users engaging just 1 or 2 days each week? Are there certain teams or functions within an organization that are getting more value, and how can I build out features to capture the teams with less engagement? Or, if the product is really driving a lot of value for specific departments — how can I understand their needs better and make sure we continue building in a direction that supports their daily workflow (and that we can upsell new features)?
The trend of over time can show if the product is getting more engaging over time
Plotting the Power User Curve for different WAU or MAU cohorts can also be very insightful. Over time, you can see if more of your user base are becoming power users, by seeing the shift towards higher-frequency engagement.
Here’s an example:
The Power User Curve for MAU cohorts from August through November shows a positive shift in user engagement, where a larger segment of the population is becoming active on a daily basis, and there’s more of a smile curve.
You can see when the line starts to inflect in order to see when a critical product release or marketing effort might have started to bend the curve. This might be a place to double down, to increase engagement. For a network effects product, you might expect to see newer cohorts gradually improve as you achieve network density/liquidity.
On an ongoing basis, you can measure the success of product changes or new releases by looking at different cohorts’ Power User Curves. If a product unblocks a bunch of features for power users, you might see a gradual increase in power users.
The Power User Curve can be based on core activity, not just app opens or logins
The frequency histogram can be keyed on actions beyond the visit — did someone show up or not — you can also go with deeper user actions. For instance, you may want to plot the core activity that maps closely to how your business is monetized… or that better represents whether users are getting value from your product. This is important because it forces you to think about what really matters to measure.
The above chart for a content publishing platform shows the total number of days in the month users posted content. A lot of products have smile-shaped core activity Power User Curves, because while most people tend to contribute lightly, there is a small contingent of users who are power users. Think of the distribution of Youtube creators, or Ebay sellers, or even how often you post on Facebook.
As the CEO or product owner of a platform like this, it’s important to design the platform such that the everyone has a chance to succeed. On Facebook, the news feed algorithm makes sure that if you feel strong affinity to a person or organization, you’ll still see their posts even if the sheer volume of other content (for instance, from more prolific media companies) would otherwise drown it out. On OfferUp, even if I seldom sell items, when I do list something, their algorithm makes sure that it’s surfaced to the relevant potential buyers.
Why does this all matter?
Not everything is a daily use product, and that’s okay.
Power user analysis allows you to get a better understanding of how users are engaging with your product, and make more informed decisions using that data. That might mean choosing an appropriate business model that works for your pattern of engagement, or designing better re-engagement loops for lower-engaged user segments, or doubling down on use cases that your high-engagement user base is already getting value out of.
The beauty of the Power User Curve over DAU/MAU is that it shows heterogeneity among your user base, reflecting the nuances of different user segments (and therefore what drives each of those segments). Creating versions of Power User Curve by various user segments can also be particularly insightful. For instance, for a business with local network effects (like Uber or Thumbtack), showing Power User Curves by market can reveal which geographies are developing density and strong network effects.
Power User Curves show if your product is hitting a nerve among a super engaged core group of users, even if perhaps the overall blended DAU/MAU is low. It also doesn’t have to just reflect app opens or logins — you can hone in on an action that maps closely to users getting specific value out of your specific product and plot the Power User Curve for that action. The key for founders is to know that there isn’t a single silver bullet to measure perfect engagement — rather, the goal is to find the set of metrics that are appropriate for their businesses. Comparing the Power User Curve of a social app vs. a work collaboration app doesn’t make sense, but looking at your own Power User Curve over time, or finding benchmarks for your product category, can tell you what’s working… and what’s not.
1/ Consumers increasingly trust Amazon, Stitchfix, and other aggregators to sort and surface the best products, which diminishes the importance of individual manufacturers. This means the algorithm is the new consumer brand–signaling trust and quality.
2/ Brands used to play an important role in an offline world with limited access to information, acting as a proxy for preference, functionality, and quality. For instance: I need to buy detergent, and I trust that Tide will do the job well.
3/ But with ecommerce, algorithms incorporated user preferences, reviews, and product traits, and became smarter with more purchases & data. “Amazon’s Choice” or “recommended for you” fill the same need that brands used to, giving us confidence that we will like something.
4/ The decline in importance of brand is not limited to utilitarian, Amazon Basics-able categories. It’s also happening in categories where brand used to really matter.
5/ In fashion, personal styling companies like Stitch Fix capture data on users’ style and size, and bypass user choice to send them clothing they’re likely to love. If you ask a customer where an item is from, she’ll likely say “Stitch Fix,” not the actual underlying brand.
6/ In the travel vertical, Airbnb and other aggregators have also eroded the power of hotel brands. Hotels represented a certain experience and quality–the name was so important that most public hotel companies’ revenues come from franchising.
7/ But user reviews and algorithmic recommendations chipped away at this brand advantage, enabling independent operators to compete on the same footing as established hotel brands.
8/ As more verticals get “Stitch Fix’ed” by shopping platforms with data network effects and growing user trust, the winners are long-tail brands that can be discovered and build healthy businesses. But the losers are big brands who can’t win on the basis of brand alone.
9/ So what does this mean for founders who are aiming to build consumer companies?
10/ First, the platform always wins. If you rely disproportionately on digital platforms at any point in the purchasing funnel, it’s important to realize that your ability to reach consumers depends on a black-box algorithm.
11/ Secondly, not all consumer categories are created equal. Some are less prone to algorithmic disruption than others. In some categories, brand will remain important for years to come. More attractive categories for brand-building include:
12/ a) Categories where look and feel and emotion matters just as much, or more, as the underlying function. No one will ever say to Alexa, “Buy a smartphone” because they care too much about the intangibles of look and feel.
13/ Many women are happy to pay an extra $30 for a Dior mascara even though there’s widely known drugstore dupes, simply because of how it makes them feel.
14/ b) Categories where people’s taste is hyper-unique and specific. Food is a literal example of this: there’s actually preferences at the chemical level. Being a Pittsburgh native, I will only eat Heinz ketchup because it actually tastes better than other brands!
15/ c) Categories where brand is closely tied to a sense of identity. Strong brands can create a language for self-expression, where having their product feels like gaining membership in a club.
16/ I’d wager that ridesharing and AVs will be more disruptive to lower/mid-tier car manufacturers where the product is viewed as a utility, vs. luxury brands where ownership is just as much a signal of personal identity.
17/ Outside of the core product, there’s also a universe of other factors that can serve to shore up brand defensibility–for instance, the content and community that build user loyalty and foster an emotional connection. Or, a unique shopping experience that serves as a draw.
18/ And removing yourself from the algorithm entirely can also work, if you can get enough leverage. Existing outside of aggregators like Amazon, and creating a standalone presence for your brand online and offline is a sure way to own the end customer relationship and mindshare.
19/ Overall, the winner in this shifting power balance between aggregators and brands is the consumer. We’re lucky that instead of spending time & energy sifting through products, or relying on brands as shortcuts, we now have access to services that do the hard work for us.