- No subscription service can retain 100% of users 😉 So, the goal should not be to retain the number of users, but rather the amount of money we earn from those who remain. Here’s a simple example.
- Let’s say 100 users subscribed to our service, paying $10 each for the first month at the basic tariff — totaling $1,000. After six months, 30% of these users dropped off. Retention is only 70%. Is this bad?
- However, of the remaining 70% (49 users), 70% continued to pay $10 for the basic tariff, while 30% (21 users) upgraded to the advanced tariff at $30. As a result, from the remaining 70% of users, we earn $1,120 per month instead of the initial $1,000. This turned out to be good, not bad 😉
Project Essence
“Just a little snag, and the whole bird is lost” — this only happens quickly and easily in proverbs 😉 This is something all developers of cloud services know. After all, the initial subscription to a service is just the “little snag,” after which a noticeable percentage of users don’t fully engage and easily unsubscribe from the service.
To prevent users from unsubscribing, they must start using the service fully — applying all its useful features. But first, they need to learn about these features.
Of course, they may have read about these features on the website at some point — but they’ve probably forgotten about them already. Moreover, new features are regularly added to products — which also need to be communicated to existing users of the product.
It turns out that informing users about existing and new features of a product they have already subscribed to is a separate task that stands fully before developers of cloud services.
The Resonance platform helps them solve this task. And it’s not as simple as it seems.
Firstly, users cannot be forced to read documentation — only a very small percentage of users are capable of doing so.
Secondly, constantly bombarding users with pop-ups offering to learn about some feature of the product or try it out is out of the question — it will annoy them so much that they might unsubscribe because of it.
This means that prompts about product features need to be shown intelligently:
Show them based on what the user is currently doing in the product — so that the prompt becomes an “assistant” rather than a “distraction.” The frequency of showing prompts to one user needs to be limited — to avoid causing irritation.
Moreover, prompts can be shown in different formats. The image shows examples of how such prompts look in different services.
Thus, developers are faced with an equation with three unknowns:
Which prompts are better to show at what point of product usage? How often can they be shown? In what format are they better to be shown? The platform’s AI engine handles solving this problem instead of the developer. It automatically selects the best moments, frequency, and appearance of prompts to maximize their clickability. According to the startup’s claims, the AI engine can find the optimal mode for showing prompts 7 times faster than if these experiments were conducted manually by the developer.
The website doesn’t mention this, but surely these modes can be tailored not only in general but also for each user individually. Developers themselves are unlikely to dare to do this, but the AI engine is perfectly capable of handling it.
To integrate the Resonance platform with the product, programmers need to insert platform calls on the product’s pages once.
After that, there’s no need to program anything to launch new prompts for new features. Therefore, marketers and product managers can independently edit and add these prompts on the platform.
If existing prompt formats are used, launching a new campaign to inform users about a new product feature will only take a couple of minutes to set the initial set of criteria for showing prompts — which the platform will then optimize on its own.
If new prompt formats are added, this time will increase to 10–20 minutes — which is also not critical overall.
Resonance was created last year, and it is currently going through Y Combinator, from which it received the first $500,000 in investments. The startup announced the launch of the first version of its platform less than a week ago.
What’s interesting
Efficiently informing users about product features is just a tool.
The real task that Resonance helps to solve is increasing Net Revenue Retention (NRR), also known as Net Dollar Retention (NDR).
The essence is that user retention can be evaluated in two different ways:
Logo retention — the percentage of users who continue to use the product after some time. Net Dollar Retention (NDR) — how much money we are currently receiving from users who subscribed to our service some time ago. NDR is also calculated as a percentage of the amount of money paid by all users who subscribed to our service some time ago. Here’s a simple example of calculating Logo Retention and NDR:
For example, in a certain month, 100 users subscribed to our service, paying $10 for the first month for the basic tariff. After six months, 30% of these users dropped out, and 70% continued to use the service. Then Logo Retention = 70%. However, 70% of the remaining users (49 users) continued to pay $10 for the basic tariff, and 30% of them (21 users) switched to the advanced tariff priced at $30. Then NDR = 100%(4910 + 2130)/(10010) = 112%. It turns out that in terms of the number of users remaining from the initial cohort, we have decreased, but in terms of the amount of money earned from the remaining users, we have increased! And what is more important for us? 😉
NDR is a much more important metric than Logo Retention because some users from our service will inevitably drop out after some time.
Therefore, the whole trick of making money in a subscription model is to get as high a percentage as possible of the remaining users to pay us as much as possible.
For good cloud services, NDR should be around 110–120%, and for especially outstanding ones, it can reach up to 150%. The image shows the best examples of NDR for cloud services that went public during the period from 2017 to 2020.
Users can pay more for increased activity, more licenses… and for a greater number of additional product features they use — either as separate payments for these features or as an upgrade to the next tariff, which includes these features.
Thus, informing users about the features of the product they have already subscribed to turns from a marketing whim into a critical task that every developer of a cloud service operating on a subscription model needs to solve. And that’s exactly what today’s Resonance helps to do.
Since this task is important, Resonance is not alone in tackling it. Another startup tackling a similar task is Appcues. Appcues raised $52.8 million in investments, attracting new investments shortly after my review.
Where to Run
The subscription model is the basic business model that cloud services operate on.
It is usually considered that the main task of such services is to attract as many new users as possible. However, as the service saturates the market, acquiring new users becomes increasingly difficult and expensive. Moreover, the number of potential users of any service, unfortunately, is not infinite.
Furthermore, if a larger number of newly acquired users drop off, the service becomes like a leaky bucket — no matter how many new users you pour into it, there will only be water (money) at the bottom.
Therefore, the very first task of any subscription service is to ensure user retention. But no one can retain 100% of users. Therefore, the primary goal is to retain not the number of users, but the amount of money earned from them, i.e., NDR.
The first critical moment for this is the creation of a pricing system in which more active and/or advanced users can pay more for greater activity and/or a broader set of product features. Without such a pricing system, having only one fixed tariff, it will be difficult to grow well.
So, the first conclusion is to more actively develop and implement such pricing methods into your own subscription-based services.
The second critical point is the implementation of an effective system for converting inactive and non-advanced users into active and advanced ones. Without this, any clever pricing system will remain just a theoretical construct.
And here you either need to start using third-party platforms like Resonance, Appcues, or similar ones… or decide to create your own platform to solve this problem and then sell it to others 😉
Considering that such a platform is actually needed by every developer of a cloud service, the potential market is very large. Well, the fact that AI will be greatly useful for the efficient optimization of such platforms makes platforms built on a principle similar to today’s Resonance very timely and relevant.
About the Company
Resonance
Website: useresonance.com
Last Round: $500K, 01.12.2023
Total Investments: $500K, Rounds: 1