How Vimeo Scaled Self-Serve Analytics to Make Data Accessible to Everyone

Insight/Action/Outcome: Vimeo identified the first experience completion rate as a major barrier to product-led growth. After introducing a new video creation flow, Vimeo noticed, through Amplitude Analytics, that the time from starting the editing process to completion got longer than expected. Amplitude Analytics focused us to identify the pain point instantly and to react quickly. By switching to a lower resolution during the real time editing stage, Vimeo increased the completion rate by 33%, leading to greater retention.

Every company wants to be data-driven, but there are just two options to get there: hire countless more people or rely on self-service analytics. In a tight market, there aren’t always the resources to hire an endless stream of analysts. That means a company has to consume data quickly—without depending on a large analytics team—to reach their data-driven goals.

Vimeo Product Org is moving towards self-service analytics, which helps stakeholders feel confident using data to make daily decisions, removing the burden from our analytics team and allowing us to move faster.

Vimeo is an all-in-one video platform that gives millions of users worldwide everything they need to make and market impactful videos. Upwards of 350,000 videos are uploaded to Vimeo daily from over 260 million active users, including 1.7 million paid subscribers, in over 190 countries. Our enterprise clients use Vimeo for internal communications, such as high-quality broadcasts of live events, video training and personal messaging, content management, and team collaboration. Our platform also supports marketing teams. They often leverage Vimeo to create video ads, which many small businesses use to create and host ads for social media. In addition, we have a sharing platform where users can embed our high-quality player within their websites, and we provide them with viewer and engagement analytics.

I’m on the product analytics team, which supports the organization by defining, measuring, and tracking our North Star metrics. We engage with the product, working closely with PMs, engineers, and designers to offer insight into user behavior. I developed the analytics team in Israel after Vimeo acquired my previous company, Magisto, which designed the video creation and editing tool. The launch of the creation and editing tool was the POC Vimeo did with Amplitude Analytics, which I led on the product and analytics side.

Removing the bottleneck from the analytics team

Answering a question about a funnel is a simple enough task for any analyst, but before Analytics, our team had become a bottleneck. Product managers with questions about specific user journeys had to go to the analytics team to build the required dashboards. The task was added to our backlog and when got selected for execution, our team would then write custom SQL, and while the task might take only a few hours, the backlog of questions started to pile up.

Addressing these day-to-day questions meant the analytics team could never handle the high-level responsibilities that required more in-depth analytics skills. And some departments never got a chance to explore the data at all. Every move required advance planning and custom work, leaving no room for spontaneous questioning.

Vimeo brought in Analytics after having used Mixpanel for nine years.

I remember being in a meeting shortly after we started using Analytics—we had just launched the creation and editing tool, and this meeting involved the product leaders, our COO, and many other people who were closely monitoring the launch. During the meeting, people began to kick around ideas and changes we could make to a particular funnel. I told everyone to hold their ideas for a minute and look at my screen—we could examine the funnel in Analytics and eliminate all speculation. Within minutes, our questions were answered.

For attendees, this demonstration was mind-blowing. Answering questions like this used to take something between hours to days —an action item for the next meeting. People said, “You just did that in real time?” Clearly, we were onto something.

Using dashboards to put data in everyone’s hands

Before long, multiple teams understood the value of this self-serve tool that allows non-analysts to engage with data without any coding. Manually building dashboards can take days. Analytics saves our data team time by enabling non-analysts to click through these dashboards and gain insights.

The funnel chart, which allows users to track the user journey within the product, is our most popular. Through it, our product team can identify the areas with the biggest drop-offs, then drill down to perform a more thorough analysis in that specific area. The funnel dashboard also shows the time to convert and the last action the user performs before converting/dropping, within a click.

Another popular dashboard is the segmentation chart, where we can track the volume of usage, either in real time or in a given time frame, such as hourly. We monitor this one closely when we launch new features and want to see early signs of adoption.

We’ve also increased our use of the retention chart, which shows what share of users returned to perform an event within a specified period. The product team might want to look at the retention between a user uploading a video and uploading a second video, for example. It used to be that if the product team wanted to change the events in the retention chart—let’s say they now wanted to look at retention of distributing a video, instead of uploading it, for example—I would have to code that for them manually. With Amplitude dashboards, the product team has the flexibility to select the events they want to see. That reduces the burden for the analytics team and allows the product team to experiment as new ideas arise.

Pinpointing time extension to improve conversion

One of our most popular use cases for Amplitude is monitoring time between steps of the funnel, a longer process usually indicates complexity or loading times. Growth is more than having a great product and flashy features. It's about efficient implementation and a fluid user experience. A frustrated user will likely find an alternative if things are not intuitive or slow to load.

Analytics has helped us track times by showing the median time to convert from one step to another, over time. Writing this is very complex in SQL, but we can identify those pain points with a click in Analytics.

We wanted to track times with the launch of our new video creation flow. Previously, Vimeo had two creation flows: one from a template and the other, through our quick video maker that took customers through 3 quick steps, guiding our AI to create their video. The first step was to select media, and as people moved through the process, the platform uploaded the media at the back. By the end of the flow, the video was already available.

Our third creation flow, Start From Scratch, is a blank editor that allows users to create a video on a traditional editor. This feature was the biggest draw, but had a lower completion rate. Looking at the time to convert in Analytics, we saw that users are waiting longer between uploading their media to the editor and performing editing actions

We addressed this by uploading a lower resolution of all media for the initial editing stage, to allow real time editing, while the high resolution uploading in the back. By reducing the uploading time for media, we shorten back the times and we increased our completion rate by 33%. We also noticed that users are more satisfied with their videos from this flow compared to the other two.

Our analysts can focus where they’re genuinely needed

Enabling self-service allows Vimeo to do more. I think about how it used to take me three hours to create a funnel analysis manually, whereas now it’s available in an instant. When we can save three hours with a couple of clicks in Analytics, that’s a game-changer for our day-to-day work.

Vimeo is a global company with over 1,200 employees, and we will need to consolidate our data structure more to encourage self-serve analytics across the entire organization. But Analytics has already saved us time and reduced the burden on analysts, and we are starting to expand its use with other Vimeo teams, like engineering and sales. Instead of generating reports for these teams the way we used to, we educate them on analytics as we guide them through the dashboard-building process.

For the data team, this will mean more time to be proactive and innovative, and to focus on that top layer of responsibilities we could never address. The increased independence of other teams means we can perform more research and prediction modeling—tasks requiring more complex analytics skills. For Vimeo, adopting Analytics means we can scale data to grow the right way: enabling PMs to be more efficient and make product decisions based on data.