Want higher NRR? Improve your Activation Rate

Net revenue retention (“NRR”) is the most important SaaS metric. It measures the year-over-year growth in ARR from existing customers; after accounting for churn, downgrades, and upgrades.

An NRR of 120% means ARR from existing customers grew by 20%. NRR is a built-in growth engine. For instance, Snowflake IPO’d at $500m in ARR. With an NRR of 158%, Snowflake could double to $1b in ARR in a little over a year, without adding a single new customer! This power of compounding is why NRR is the most important SaaS metric.

Despite its importance, NRR is a difficult metric to improve directly because it is a “downstream metric” which is the result of many actions that happen months in advance of upsell or churn. So, if your company has low NRR the best way to improve it is to find an upstream metric that is easier to move and has positive covariance with NRR.

In this post, I present the best upstream metric for NRR: “activation rate”. Here’s a quick summary of this post:

• Activation is the minimum activity that shows the customer is “deployed”.
• NRR decays with activation time.
• Customers that never activate almost always churn.
• You should set a goal for your product & customer success team to lift the activation rate year-over-year.

Now, the details:

What is activation?

Activation is the minimum activity that shows the customer is “deployed”. Deployment means the customer understands how to use the product and has experienced some of its value. If defined right, an activated customer is less likely to churn than an inactive one. That is why the activation rate (% of customers that activated) predicts churn & NRR.

Let’s look at a few examples to understand the point:

• LinkedIn defines activation as 30 connection in 30 days
• Facebook: 10 friends in 7 days
• Invision: Create and share 1 project OR share 1 project and leave 1 comment, in 30 days

A few patterns emerge from these examples. A good activation definition is:

• Simple. It involves a few variables that are all fundamental building blocks of the product: friends, users, projects, etc.
• Memorable. “30 connections in 25 days” is more accurate, but it is difficult to remember than “30 connections in 30 days”. A memorable definition rallies a team behind a goal because you cannot move what you can’t remember.
• Time-sensitive. The quicker a customer activates the less likely they are to churn. The exact time horizon increases with product complexity: it is easier to add 10 friends on Facebook than to create a project in Invision.
• Persona Dependent. LinkedIn has a different definition for recruiters who do not need connections to get value from LinkedIn. So, if you have multiple key personas (e.g. a marketplace), then build a definition for each persona.
• Balanced. Getting activated should minimize churn. However, it is possible to set the activation bar so high that churn is practically zero (e.g. 100 friends on day 1 for Facebook) but the activation rate — the percentage of customers that activate — is also zero. Such a definition is useless. A good definition maximizes activation rate while minimizing churn rate.

Let’s exercise these patterns and create activation definitions for a few other companies:

• Zoom: 100 minutes of video calls in 5 days.
• Slack: 5 engaged users and 2 channels in 7 days
• Salesforce: 3 users and 10 opportunities in 14 days

You get the point.

Activation time predicts NRR

Why is activation important to NRR? Because it is an upstream metric to NRR that:

• Predicts NRR
• Is easier to improve than NRR
• Is an early warning signal, as activation happens within the first few days to months of customer life.

Because of these reasons, if you want to improve NRR, improve your activation rate.

Let’s see how activation time predicts NRR:

The following graph is synthetic; it shows how Split’s NRR decays with activation time without revealing its NRR or activation timeline. The data isn’t real, but the lessons are.

NRR decays with an increase in activation time, irrespective of the segment. This makes sense; a quick activation can mean a few things working in favor of expansion:

• Empowered Champion: You have a strong champion at the customer who can marshal the resources necessary to deploy your product.
• Urgency: She also has a “compelling event”; if she doesn’t activate her business will suffer in some form.
• Product-Problem Fit: Your product perfectly fits her problems. You do not need to build another feature to unblock her.

Together, these factors lead to a quick activation, which means the customer experiences your product’s value quicker. A customer that experiences value will expand over time.

Enterprise NRR decays slower than SMB/Mid-Market

Enterprises are slower to activate than other segments. As a result, their NRR doesn’t decay as much as SMBs and Mid-Markets. In other words, there is a grace period built into enterprise activation.

This is because SMB champions have a lot on their plate and little time; if they can’t get activated on the product quickly, they will move on to the next priority or even switch to a competitor to get success. Enterprise champions, on the other hand, are used to slow implementation cycles. They have to jump over enough hurdles within their own company, even after the sale, that they are forgiving to a slower activation.

Enterprises are not forgiving of non-activation

SMBs and some mid-markets may never activate if the activation definition is dependent on team size. In such cases, the customer may be perfectly happy and renew year-over-year.

However, enterprises that do not activate do not renew. That’s why it is important to understand activation so well that you can predict which enterprise customer will take how long to activate so you can insist on multi-year contracts with them to allow for unexpected slowdowns on the path to activation and expansion.

What is a good activation rate?

As a rule of thumb, you should activate 90% of your customers in the shortest practical time. The shortest practical time depends on the company. For Facebook / LinkedIn, it’s in days, for Salesforce it’s in weeks. The exact time doesn’t matter. What matters is pushing your product and customer success team to lift the activation rate every year (or quarter) so you get a graph like this:

This hypothetical team is continuously pushing the activation rate up and early, which will lift the company’s NRR.

Final Thoughts

Thank you for reading. I will cover more activation-related topics in future posts. In the meanwhile, I’d love to hear how you define activation and whether your activation definition predicts NRR.