Insight/Action/Outcome: One product team encountered a 45% dropoff between sign-up and setup in their app, partly attributable to a particular module that users quickly left after entering. Further investigation showed users were confused about how to engage. Experiments with a new call to action are ongoing, but similar Appfire experiments to remove friction have resulted in a 10–14% increase in post-experiment product activation metrics.
Any company serious about adopting a product-led growth (PLG) strategy needs rock-solid data. They also need a tool to make that data trustworthy, accessible, and understandable. For Appfire, it was more than a tool.
Appfire is the leading provider of software that enhances, extends, and connects the world’s leading platforms, such as Atlassian, Microsoft, monday.com, and Salesforce, enabling teams to thrive and do their best work together. Our solutions help more than one million users with everything from DevOps and IT service management to business intelligence, product portfolio management, automation, collaboration, and more.
As Director of Product Growth, my latest challenge was building a complete product analytics program where we had no formal program before. We had to find the tools and design the taxonomy and workflow for bringing apps and teams on board. This task is easier said than done.
Because some of our apps and products came to Appfire through acquisitions, certain product teams had their own cultures, tech stacks, and methods of development. Each of these teams brought their unique strengths to Appfire and were at various levels of maturity when they joined us, with some bringing their own product analytics (or not). We had to be flexible enough to work with the particulars of each team yet rigorous enough to establish a common standard of data quality that would engender trust in the data.
Empowering leadership to find common threads among many apps—and to trust those insights—required applying a common taxonomy.
It was a big task, but a consistent data program was necessary. As a portfolio company of 100+ apps, we need to understand not only each app individually but the entire portfolio more broadly. Empowering leadership to find common threads among so many apps—and to trust those insights—required applying a common taxonomy. This taxonomy would need to be applicable across our existing ecosystems and their different nomenclatures, as well as any future acquisitions.
The only platform reliable and trustworthy at scale
My team wanted to avoid the common mistakes we often see in our industry. Half of what you read about product analytics involves advice for fixing analytics and making data usable. Rather than fix data quality and other issues, why not adopt best practices from the beginning? We needed a platform that supported that level of rigor and would enable us to establish taxonomy standards and monitor and review those standards to help teams adhere to them. We wanted to create a portfolio activation dashboard showing performance across all instrumented apps. To perform that kind of apples-to-apples comparison, we needed the data governance capabilities that would maintain that consistency.
Half of what you read about product analytics involves advice for fixing analytics and making data usable. Rather than fix data quality and other issues, why not adopt best practices from the beginning?
Our analytics also needed to go beyond just capturing page views and events; we needed to understand behavior. And because we’re a B2B company, we needed to understand our customers in multiple dimensions, including capturing the user journey at the account level. So many platforms lack the level of maturity to handle B2B or the scale of our operations. They’d require us to make many compromises, and my team would still have to do a lot of manual work to hit our high standards.
Amplitude’s data governance capabilities make data reliable and trustworthy at scale. It was the best option for the project.
Selling teams on what they could accomplish
My team did a lot of preliminary planning to get everyone on the same page. For the first three months after selecting Amplitude, we ran a pilot with our most sophisticated app and our most demanding team so we could discover our biggest challenges upfront.
From there, we created a documented plan for instrumentation. We tried to document as much as possible and make it easier for product managers, developers, the UI team, and everyone else involved to find information about the new framework.
We wanted to approach this process from one of Appfire’s core values, “Getting it right > being right,” where more feedback leads to clearer understanding, greater focus, and more improvement. For us, that meant working with teams to overcome the hurdles they saw rather than imposing an inflexible set of rules, even as a high standard of data quality remained our top priority. Engineers have a hard job, so we wanted to do everything possible to make it easier.
Engineers have a hard job, so we wanted to do everything possible to make it easier.
We also educated app teams on PLG concepts, giving them a vision of what these analytics would enable them to accomplish. Without exception, the enterprise edition of Amplitude was superior to any analytics platforms other teams were using. Some teams had been using the free version of Amplitude, so the enterprise edition was a welcome upgrade.
We acknowledged there would be some work in adopting our common taxonomy, but this taxonomy, coupled with enterprise analytics, offered incredible capabilities that would be worth their effort. It wasn’t hard to sell what we were offering, and with that, we got buy-in and laid a solid foundation for our PLG strategy.
Helping product teams find their aha moment
We tightly integrated our instrumentation with our activation methodology so we could clearly see the results of activation experiments and improvements to the UX in our apps.
One of our products is a project management add-on for Jira. A problem arose when the product team found a 45% dropoff between sign-up and setup.
The team responded by running an experiment that improved the onboarding experience. Although this experiment is ongoing, the results are trending positively. Other teams that have run similar experiments to remove activation friction have seen a 10–14% increase in post-experiment product activation metrics.
Real-time analytics make experiments more effective
There are many ways to view success. For my team, we set out with a list of dozens of high-priority apps to target for instrumentation, and we instrumented 35 of them within the first year of our new taxonomy. With few exceptions, we successfully instrumented all the apps identified as the highest revenue contributors or otherwise priorities for investment.
In each case, instrumentation was a deliberate process of growth and product managers (PMs) working together to map the activation journey and identify the critical events and areas for improvement for that specific app. All of these apps have started making these PLG motions, with 20+ apps already running experiments to iterate improvements to their activation journey. Of these 20+ apps, all have seen some improvement in activation rates after experimentation, with the results still rolling in.
Because we have Amplitude, we can show that a 10% improvement in activation generates a 2.5–3% improvement in license conversion. All our apps are available for a free evaluation period, and our objective across the board is to activate the user to purchase before the end of that free evaluation period.
Relying on lagging indicators hinders product teams from iterating effectively. Accessing the data through Amplitude enables us to see real-time behavioral changes, proving the relationship between user activation improvements and increased conversions.
It previously took us 60–75 days to know whether someone who started an evaluation bought the product. Relying on lagging indicators like this hinders product teams from iterating effectively because they never have the timely feedback they need. Accessing the data through Amplitude, we can now see real-time behavioral changes, proving the relationship between user activation improvements and increased conversions and showing PMs how their actions affect revenue.
Changing the conversation around product success
Amplitude use is steadily broadening across Appfire, with the product growth team, product managers, and product engineering leads being the heaviest users. Building a data culture was one of my team’s critical goals, so seeing our product teams engage with Amplitude to answer their questions confirms our success, especially when some of our smaller acquired products used to lead with an engineering mindset.
We are still only in the middle of this company-wide maturation. My team is already fielding requests for specific analytics and access to Amplitude. Our product marketing team, for example, has become very engaged, and our customer success team sees potential for receiving behavioral signals that identify accounts having difficulty with adoption. This information would enable the team to offer these accounts extra support.
Now that we have behavioral data, we can describe product performance as we never could before, tying it directly to license conversion and revenue. This shifts the conversation from finance and sales to product metrics and firmly establishes the value of Amplitude as being fundamental to the product operating system.
With Amplitude’s help, we are enacting a PLG mindset at Appfire by changing the conversation throughout the organization. Not only has the executive team become high-level consumers of the data, but we see them acknowledge PLG metrics as the superior way to understand product performance.
Until we started using Amplitude, we could only say that we sold more licenses in one month compared to another. Why did we sell more licenses? No one knew. Now that we have the behavioral data, we can describe product performance as we never could before, tying it directly to license conversion and revenue. This shifts the conversation from finance and sales to product metrics and firmly establishes the value of Amplitude as being fundamental to the product operating system.
Have more questions for Michael and the Appfire team? Chat with them and other Amplitude users in our Slack community. Use this space to learn how others are using Amplitude and share best practices.
