October 01, 2014A number of our users were way over their plan limits. Here's what we learned when we pointed this out.
A few months back, we realized a large number of our users had outgrown their current plans. Some of them were several times over their plan limits - tracking well over a million page views per month on a plan that allowed for 100,000 views, for example! Naturally, we found this concerning. Not only were these users not paying for the extra views, but capacity planning was a lot more difficult with so many users sending way more data than their plans allowed.
The thing is, our users weren’t abusing Gauges by any means. Many of them had absolutely no idea they were so far over their plan limits - because we weren’t telling them. We just weren’t exposing how many of their monthly visits they were actually sending. It shouldn’t be up to our users to track their own usage. We have that information - it was simply a matter of making it available to our users.
It was hard for overage users to tell when they were over their limits, because we don’t have hard caps on web traffic. A lot of development shops use Gauges to provide easy to understand web analytics to their clients, and we didn’t want to simply stop tracking visits to their clients’ sites. On top of that, it’s often difficult to predict when you’re going to get a big traffic spike, and we don’t want to stop tracking your visitors when you’ve exceeded your limit for the month.
Still, we had to make a change. We weren’t exposing usage, and nothing out of the ordinary happened when a user reached his monthly limit. We want all of our users to be aware when they had exceeded their plan limits, and we wanted to move overage users to more appropriate plans. Our thinking was if we showed users who were over their plan just how far they were over their plan, and gently encouraged them to upgrade, we’d see more upgrades.
We started by identifying users who were at least 20% over their plan limits by making a new segment in Intercom:
Now we just needed to figure out how to persuade overage users to move to the right plan for their actual use case. We quietly rolled out a notice that appeared on the dashboard of any user who was more than 20% over their plan:
We noticed the impact of this update almost immediately.
First, we started hearing from long time users who were confused and thought we were going to shut down their accounts unless they upgraded their plans. This was never our intention - we don’t think any of our users were actively trying to abuse Gauges. We’d never exposed our user’s actual plan usage in this way before, making it harder than necessary for any of our users to remember if they were within their plan limits.
We also saw a spike in user cancellation. Shortly after we rolled out this notification, cancellations increased. Not as many as we thought we would see, but it happened. User churn was a bit higher than normal following this change, but quickly returned to normal levels.
User upgrades also became more frequent. Previously we’d seen upgrades from Solo to Small plans most often, for users who wanted to share web traffic stats with other users. Now we began seeing upgrades based on how much of a plan a user was using.
Finally, we heard from a number of users who thought this was a great idea, and were happy to upgrade their plans. Users in this camp tended to be longtime fans of Gauges, likeminded companies, and other passionate users.
We absolutely had a messaging problem around the rollout of this feature. We could have mitigated this by announcing the feature update, or making it absolutely clear we weren’t going to cancel anyone’s plan. We’re still using the same integration to expose overage, and this could be made more useful. We could switch to a running total that resets monthly, for example, rather than simply notifying users when they’ve exceeded their plan limits. We’re still seeing a number of users with overage, but this number has been steadily decreasing since we implemented this change.
We also got a better understanding of how much our users like Gauges - businesses that rely on us to track their web traffic stats were more than happy to upgrade, and more casual users simply canceled their plans.
We hate seeing cancellations - anyone who runs a SaaS business knows that your business lives or dies based on user churn. But while we did see a spike in user cancellations, this trend didn’t continue and quickly returned to normal levels. By contrast, users are more likely to upgrade once they’ve hit their monthly limit, and overall we consider this experiment to be a success.
Could we have handled this more elegantly? Absolutely. But we’ve learned a lot - about how our users actually use Gauges, and whether we could actually see a revenue bump from our active user base.