Self-directed Insights Lead to High-Quality Growth at Coursera

Insights/Action/Outcome: Noticing a spike in metrics around learners’ growth, Coursera used the Anomaly + Forecast feature in Amplitude Analytics to determine the cause: Their platform’s calls to action (CTAs) were driving more traffic. Understanding each CTAs’ importance, Coursera further optimized the most effective CTAs with Amplitude Experiment, causing 30% growth in high-value traffic and course registrations.

Just as you wouldn’t fly a plane without a dashboard, you can’t expect teams to make intelligent business decisions without having a dashboard to monitor the user experience. But that’s what we were doing at Coursera. Flying blind was holding us back and preventing us from moving ahead as quickly as we could.

As a senior director, I lead the strategy and operation functions for the marketing organization at Coursera, an online education platform that democratizes learning for anyone and everyone across the globe. Our highly curated array of classes delivers the best education and training from leading universities and companies through video tutorials.

When I arrived at Coursera, the company used a home-grown analytics system that could answer basic questions, but it wasn’t very insightful and couldn’t scale. I wanted to do more than answer basic questions—I wanted to kick our marketing efforts into high gear. Thankfully, Amplitude Analytics was part of our tech stack.

The move to democratize data and reduce time to insight

Even though Analytics wasn’t highly used by the marketing organization, I immediately latched onto the platform because I knew it held the key to moving marketing forward.

More than just analytics and visualizations, we could use Analytics to activate various campaigns across outbound channels—a unique value proposition. We could also examine different user properties and slice and dice information on the fly to make better business decisions. I took the lead on the secondary rollout efforts, and we’ve onboarded 60 people through a combination of office hours and using Analytics to solve business problems.

Our primary objective in the marketing department is to further growth and revenue. To do that, we need to democratize data and reduce time to insight. The latter is essential to us. We used to have to take business questions to other parts of the organization to understand what’s working in marketing. Now, marketers can use the information in Analytics independently, to quickly determine what’s working and not working in their respective campaigns.

Translating insight into a 30% growth in traffic

A few weeks into using Amplitude, we had our first big “aha” moment when we noticed specific metrics spiking around learners’ growth. Using the Anomaly + Forecast feature, it took just two minutes to decipher the reason behind these spikes: Certain features—calls to action (CTAs)—were driving that traffic.

Previously, we would have gone to the data science team with questions about these spikes, and waited days or even weeks for an answer. Explaining the need to another team takes time, and they’re often busy with multiple requests. But with information at their fingertips, marketers can answer these same questions in seconds.

We worked to make the CTAs driving traffic even more prominent. Using Amplitude Experiment, we iterated on the CTAs to optimize further, made their messages resonate even more. As a result, we saw traffic grow by about 30%—just from optimizing those critical CTAs.

We’ve seen a knock-on effect, too. Alongside our traffic increase, we saw a spike in registrations—specifically high-quality registrations. We don’t want just anybody to click through and flood our database with low-quality leads. We want users who are genuinely interested in Coursera’s offerings, so we refined our CTAs to ensure they were clear in what we asked people to do. With the growth in high-quality traffic to those pages, we have seen an uptick in high-quality registrants, and that’s what we value most.

Encouraging marketers to follow their self-serve data journey

The power of Amplitude is the ability to answer one question, which leads to another question, which leads to more questions. Within moments, marketers can make this self-serve journey and find the direction they should take. That democratization empowers marketers to do what I think of as “shallow fishing” for insights. We still call on our data science team for “deep fishing,” those long-term problems we need to address more strategically. But marketers know their business problems very well, and tools like Analytics and Experiment enable marketers to test their hypotheses better than anyone else in the organization.

Amplitude can solve anything, and it’s straightforward to learn: A person can train and start using Amplitude within an afternoon. At the same time, there’s always room for improvement and room to learn how to ask intelligent questions that move the needle for your business. Developing that skill is critical to marketing success, and it’s what I have challenged my team to do.

We’re amplifying our use of Amplitude in two ways by:

1. Sharing our learnings and outcomes with the leadership team
2. Continually giving people access to the platform so they can test it out themselves

We want to get this tool to every marketer in the organization, and I expect to have more than 100 users in the next few months. We are an international organization, so we see massive potential in sharing information with global teams. It’s easy to share dashboards with a colleague halfway around the world, and from there, teams can begin solving regional use cases.

Marketing doesn’t need to be perfect, but we need to be fast. Coursera exists in a very competitive space, so if we see an opportunity, we must seize it quickly. Without Analytics and Experiment, our actions would be much more delayed, and our decision-making would slow significantly. The faster we can convert learners, the faster we can democratize learning for everyone across the globe.