It’s no secret that dApp metrics suck.
Commonly used metrics like transaction count and volume aren’t necessarily representative of the performance of a dApp. Further, there are often serious data integrity concerns around the data itself.
As the Ethereum ecosystem enters the next wave of dApp development, it’s beneficial to examine a potential unified metrics framework for dApps.
By measuring user behavior metrics, dApps can optimize, iterate, and grow. And by measuring investor-centric metrics, dApps can position themselves for future funding to aid growth.
Asset vs. Application Level Metrics
Asset metrics convey the relative strength of a network or usage of an asset.
Max Mersch of Fabric Ventures does an excellent job of outlining top asset metrics across Bitcoin, Ethereum, and the Maker DAO ecosystem in his article An Overview of Relevant Metrics in Web 3.0.
Asset level metrics like Total Locked Value (TVL), for example, provide a glimpse at ETH usage in DeFi. The argument is that asset level metrics are more representative compared to trying to force traditional metrics like Daily Active Users (DAU).
How many people check in on their MakerDAO CDPs every day but do not perform a CDP-related action?
In the traditional metrics world, these users would be considered daily active users. But from an asset level, this metric does not necessarily indicate asset or protocol usage.
Contextualizing App Level Metrics
Any application built on top of the Ethereum will be inextricably tied to it’s higher order asset level metrics.
If you’re building a DeFi application, you care about value locked up in DeFi. If you’re building an application for developers, you’ll be concerned about number of developers using Truffle/Ganache/Zeppelin/Other Libraries. And so on.
And, the actual asset your application utilizes affects your growth as well.
If your dApp uses DAI, it’s likely more transactional. If your dApp uses ETH, perhaps it’s more usage focused or speculative.
It’s within this broader context that we can examine application level metrics and a unified framework that is applicable to ALL dApps.
DApp Unified Metrics Framework (DUMF)
There are many dApp uses cases — from DeFi to games to marketplaces to exchanges and beyond. Each category will have use case specific metrics that help measure and demonstrate success.
One thing all dApps, regardless of category, have in common is a need for user acquisition and sustainable funding.
The DApp Unified Metrics Framework (DUMF) serves as a simple standard framework that can be applied to any dApp to aid in both.
The framework can serve as an accessible, simple starting point for dApps — with only eight metrics total to measure.
When combined with asset level metrics, they become even more potent as a unique representation of a dApp’s use case.
Let’s dive in.
User Behavior Metrics
Tracking user behavior metrics is critical to understanding who is using your dApp, why they’re using it, and exactly how they utilize its functionality.
The top performing dApps will utilize these metrics to iterate on the user experience and continuously improve the dApp.
Onboarding Conversion Rate
Onboarding conversion rate measures the percentage of dApp visitors who successfully perform one action in your dApp.
The user funnel for dApps is much more complicated than that of a web 2.0 application. In the web 2.0 world, it often consists of visiting an application, signing up, then performing a necessary action.
In the web 3.0 world, it currently consists of visiting a dApp, installing MetaMask, acquiring ETH, and then performing a necessary action. Comparatively, there is obviously much more friction baked into the Web3 user journey.
What do you think the average dApp onboarding conversion rate currently is?
Increasingly, however, we’re seeing web3 components and providers that reduce this friction for onboarding new users. Consider an alternative optimized web 3.0 application user funnel:
This is only three steps and rivals the ease of the web 2.0 flow.
Whatever your setup is — whether that’s a MetaMask flow or a smart wallet, covering gas for your users or not — it’s necessary to measure your onboarding conversion rate.
Onboarding friction reduction = more users, happier users, and users more willing and ready to explore your dApp.
Bounce rate is the percentage of users who visit a dApp and do not navigate to any other page before leaving.
Before even performing an in-dApp action, dApp visitors will likely land on your homepage to see if they even want to use the thing.
What does the content and copy of your dApp’s homepage look like? Does it provide a compelling reason to use the dApp and properly articulate its value proposition?
Measuring bounce rate is an easy way to detect if your message is connecting or falling flat with potential users. It’s inevitably more difficult to convey the value prop of a friction inducing dApp over a blazing fast centralized alternative.
That’s why the answer to the “Why should I use this thing?” question should be abundantly clear upon arrival.
If you have a high bounce rate, think deeply about the true value proposition of your dApp. What does your dApp offer that a centralized alternative doesn’t? What is cool and novel about it? Consider altering the copy to reflect this value proposition and re-measure bounce rate (it’s an experiment!).
Beyond the value prop, bounce rate can also convey ease of use.
Think about how many times you’ve visited a dApp only to be greeted with the MetaMask connect pop-up:
Guess the bounce rate for dApps who pop up connect MetaMask upon load (seriously, guess)
Retention rate measures what percentage of users who perform an in-dApp action perform another in-dApp action within a certain number of days.
This measures the stickiness of your dApp on different time frames such as 1 day, 3 day, 7 day, 14 days, and 30 days. It does take 21 days to form a habit after all.
Do users come back to your dApp time and time again? Is your dApp addictive? Are there user rewards? Do your users perform any action at all on your dapp that forms a habit?!?!
Mechanisms like variable rewards, social interaction, or unlocking achievements can increase dApp engagement.
Consider InstaDApp’s earn dashboard. It counts your lending earnings in real time. Sure, it goes about 15 decimal places out in order to tick up every half second. But, it’s oddly addicting to look at and check back in on, right?
It’s hooks like these that will keep your users coming back to your dApp time and time again.
And if they come back to your dApp frequently to check in on earnings, it’s likely they’ll perform other, additional actions in the process.
Cohort analysis measures the extent to which various groups of users engage with your dApp.
As you can imagine, there are a variety of ways to group users — by referral social platform (did they come from Telegram or Twitter?), by demographic, by dApp usage, and so on.
Which of your user cohorts is the most actively engaged? Which is the most profitable? Which shills your dApp on Twitter and Telegram?
Measuring your most active cohorts can help you more efficiently target and acquire new users with similar characteristics.
Consider, Hxro, a gamified trading app — they discovered relatively quickly which cohort was most profitable for them — the hybrid gamer-trader.
So then they started doing Twitch streams to target this demographic:
Every dApp likely has a hybrid demographic that can be specifically targeted. Finding this cohort of die hard users is the hard part.
User behavior metrics allow you to optimize and iterate your dApp. An equally important consideration — which metrics would you show to a potential investor or grant giver to demonstrate the utility or growth potential of your dApp?
Funding sources in the web3 space are evolving. There are grant giving DAOs like MetaCartel. In the future, app DAOs may facilitate revenue sharing as an on-chain version of revenue-based financing. And, of course, there are a variety of Web3 focused VC firms.
All three entities will likely request some metrics to better understand the current usage and future potential of your dApp.
Increasingly, dApps are experimenting with and implementing sustainable business models to generate revenue.
The monetization potential of any dApp is a strong consideration for investors or DAOs seeking to disburse grants.
Aside from pure revenue generated, there are a few revenue-related metrics which provide more granular insight into exactly what is driving revenue growth — whether that’s a high revenue generated per user or perhaps a high acquisition cost.
Average Revenue Per Daily Active User (ARPDAU) — is most of your revenue coming from a few active users or have you monetized a wide base of users?
Lifetime Value to Cost of Acquisition Ratio (LTV:CAC) — signals the sustainability of revenue growth. If the ratio is too low that signals high cost of acquisition — indicating that current growth may not be sustainable unless subsidized by future funding.
DApp growth rate is defined as the percentage increase in total daily active users on a weekly basis.
This demonstrates product traction. Beyond traction, part of the growth story is a viral coefficient.
Viral Coefficient — the number of users an existing user generates. This is applicable only if there is a viral component to dApps. For games and content platforms for example, the viral coefficient is extremely important to bringing new users to the dApp.
Web 2.0 applications commonly measure Net Promoter Score (NPS).
NPS is measured with one simple question: “On a scale of 0 to 10, how likely are you to recommend to a friend?”
Anyone answering in the 0–6 range is considered a detractor, 7–8 range a passive user, and anyone in the 9 to 10 range is considered a promoter.
Crypto, and the Ethereum community specifically, heavily utilizes Twitter and Telegram. If you’ve ever stumbled upon an amazing dApp, it’s likely that you found it through someone who shilled it on Twitter or Telegram.
So, perhaps the Web3 equivalent of NPS is Twitter Promoter Score (TPS) or Telegram Promoter Score (also, TPS).
TPS could be measured with one simple question: “On a scale of 0 to 10, how likely are you to shill this dApp on Twitter or Telegram?”
It may sound silly but…
Community is extremely important for dApp adoption.
These platforms and a high promoter score in general are an excellent signal that the community is genuinely engaged with your dApp.
Test out DUMF on your dApp
If your dApp is serious about growth, sustainable funding, or both, strongly consider measuring DUMF and holding your team accountable for reaching specific metrics benchmarks.
This is a proactive experience, not just measuring data for the sake of data— but learning from the metrics, experimenting, then measuring the effectiveness of the experiment, and integrating necessary changes into your dApp.
At the very least, it’ll be an eye opening experience.
And hey, it may just position your dApp to gain more users, generate more revenue, receive that grant you’ve been applying for, or finally get investment.