You did your research. You contacted the customers. You gathered the requirements. You designed it, you built it, you launched it. You celebrated the release.
But nobody is using it.
It’s been a few days since the launch. Or maybe even a few weeks. Maybe there was a flurry of initial usage that slowly turned into a trickle, or maybe you never really got any traction at all. What gives? What happened?
Defining Your Funnel
Sometimes, we do mess up. Maybe we contacted the wrong customers, or we put together the wrong set of requirements. Maybe what we built didn’t match the requirements, or the requirements didn’t match the customer. Maybe there never was a customer. Whatever the case, you need to do something, because the product won’t adopt itself.
Diagnosing what went wrong isn’t always easy, but getting the diagnosis right is crucial to fixing the problem. One way to do this is with Dave McClure’s “AARRR” funnel:
- Customer Acquisition
- Account Activation
- User Engagement / Retention
- Revenue
- Referrals (Virality)
Each of these steps must be well-defined and easy to measure. For example, we define “Customer Acquisition” for our Traffic Platform as the instant that somebody hits our signup page. This might be different for other companies: It might be marked by a customer’s first contact with the company, or their first visit to a website, or a download. What’s important is that, whatever the event is, it can be easily described and easily tracked.
What each of these steps looks like will vary greatly depending on your business and your product. In many cases, each step may be a funnel in itself. On Traffic Platform, a user must first provide their name and contact information, then fill out a profile, and finally provide funding. Only once their account has been funded do we consider the account fully activated. Similarly, retention should be measured periodically, at fixed benchmarks. Do users continue to use the site after a day? A week? A month?
Finding the Leaks
Once you’ve collected data for each of these steps, you can start to look at how to diagnose where your product has gone wrong. Let’s imagine your funnel looks something like this:
| Step | Definition | % Retention from previous step |
| Acquisition | User lands on signup page | 100% (34,000 unique page views) |
| Activation | User funds account | 0.5% (170 funded accounts) |
| Engagement | User creates 3 campaigns | 88% (150 engaged accounts) |
| Revenue | User spends $500 | 67% (100 spending accounts) |
| Referrals | User makes first referral | 25% (25 referring accounts) |
Note: All figures and benchmarks are entirely fictitious
This example is made up, but the numbers are not entirely unrealistic. Hopefully, that 0.5% in the Activation step jumps out at you. Clearly, that’s where our funnel is leaking, and our activation process needs to be examined.
What to Do When You Jump the Gun
But let’s take a step back and imagine for a moment that we didn’t bother to inspect our funnel. Maybe we launched our product and didn’t see the usage we were expecting, so we decide to redouble our marketing efforts. That might work, but it’d be very expensive, and probably not sustainable. After all, only one in 200 of the people who hit our signup page will ever activate their account (and only one in 340 will ever hit their target spend).
Recently, we launched our new Publisher Dashboard. The new Publisher Dashboard replaced the old dashboard, making it cleaner, more consistent with our company branding, easier to use, more functional, and more performant. In our eyes, it was a clear winner across the board, and we were expecting our publishers to be chomping at the bit to make the jump.
Classic Dash usage:
New Publisher Dashboard:
What actually happened was the exact opposite: We had rocky adoption at first. We saw some publishers go over to check it out, but not nearly as many as we had expected. To our surprise, many of the publishers who did log into the new platform continued to use the old dashboard. Clearly, we were missing something. Our users were already acquired, so for this case, we focused on the Activation and Engagement steps of our funnel.
For our Publisher Dashboard, we defined Activation as simply accessing the new dashboard. To facilitate that, we added a modal and a header to our classic Publisher Dashboard inviting publishers to try to the new interface. We also asked our Sales Team to reach out to our publishers with a personal invitation to the new platform. While this did attract some attention, it didn’t win over all our publishers.
As developers, we were confused. Who wouldn’t want more drill-down reporting functionality, 10x faster query times, and a UI that made the old design look like a four-year-old’s crayon drawing?
Stepping into the users’ shoes, though, revealed another story: On the classic dash, 60% of normal sessions involved interacting with only a single page before leaving.
Our users on the old dash didn’t need a cleaner design to do what they were doing. They were hitting a single page, pulling some stats, and leaving. Moreover, our messaging didn’t indicate that the new dashboard was 10x faster, and it didn’t say anything about our new drill-downs or reporting options. These were things we knew that a few publishers wanted, but they weren’t things most publishers knew we were offering. When they saw “New Pub Dash”, they thought to themselves, “Eh, this already works well enough.”
The publishers who did investigate our new dashboard fell into two groups: Those who simply viewed the same content they were viewing on the old dash, and those who clicked around to see what other options they had. Unsurprisingly, it was the ones who viewed a single page who reverted back to the classic dash, and the ones who clicked around who stayed. In fact, current usage of the new platform shows that 70% of sessions result in interacting with three or more views, compared to only 37% on the old dash.
Effectively, the funnel informed us that our activation rates were poor, but our engagement rate was high. This was reassuring: we had built something that people wanted to use after all. We just hadn’t done a very good job telling people about it.
The Solution
Eventually, after identifying the problem and much haranguing, we did increase the usership of the new platform to the point that we felt comfortable shutting the old platform off. Since then, we’ve seen nearly a 50% increase in page views (compared to the classic dashboard), a decrease in time-on-page, and regular engagement from several hundred users who had accounts on the classic dashboard but showed sporadic or nonexistent usage. All in all, not too bad.
What does this mean for your company? While particularly useful for new products, the funnel is valuable any time you want to look for obstacles getting between your company and revenue. It can show where to allocate resources and what to prioritize. Oftentimes, it’ll reveal that bottlenecks exist in places you wouldn’t expect—like in our case, where the problem wasn’t in the platform itself, but in the messaging we used to inform users about it. And, by using the funnel, you can identify and remove those bottlenecks, resulting in happier users, higher spend, and a stronger bottom line.