This is a four-part series
Click below to navigate to the next chapter:
Chapter 1: A Nervous System of Data sits at the Center of the Modern Company
Chapter 2: Why We Need to Look Beyond CDPs to Deliver Excellent Experiences to Customers
Chapter 3: Snowplow and Census: Delivering Behavioral Data for Actionable Outcomes
Chapter 4: Enhancing the Snowplow Trial Experience With Census
Enhancing the Snowplow trial experience with Census
At Snowplow, we had a problem with product discovery. We were confident in our tooling – we knew our behavioral data engine was robust, arguably the most expansive, reliable solution for collecting rich behavioral data. The problem was (and sometimes still is) our prospects and customers didn’t know this. Worse, they had no easy ways to find out exactly how Snowplow could help them and their business. We could wax lyrical about the benefits of great data quality and rigorous processes for validation, but so what? It’s difficult to truly understand such abstract concepts without seeing them for yourself. Without experiencing the advantages of Snowplow data, all our most passionate marketing and storytelling would leave our prospects cold.
We realized we needed a way for our prospective users and customers to try Snowplow for themselves. Before you invest in an expensive car, you want to take it for a test drive, after all. The same is true for an investment in software, especially when the tool handles your company’s most important asset: your behavioral data.
We got together and launched a trial version of Snowplow, quaintly named “Try Snowplow.” To give people a taste of the software, users could sign up to a ‘Try Snowplow’ account, spin up a small version of the Snowplow pipeline in their own cloud account and fire off events into their data warehouse, all in 30 minutes or less.
Naturally, we were curious to see how the trial experience would be received. But beyond that, we wanted a deep understanding of those using the product, and how we could make it better. We wanted to know what challenges they would face, the bottlenecks they might encounter, and whether or not the experience would help them discover the benefits of the Snowplow product.
How we measure Snowplow trial engagement
In order to better understand our users, we assembled a stack to capture their interactions at each stage of the trial experience (without, of course, seeing what data they capture for themselves). In a nutshell, here’s what our data stack for the “Try Snowplow” experience looked like.
We deployed tracking across the web UI for Try Snowplow as well as Scala tracking in the Try Snowplow pipelines themselves. Snowplow sends this data into our data warehouse (Redshift), where we’re able to use SQL Runner to aggregate important pieces of information together to build a picture of our users. It also means we can qualify potential leads early in the process, before alerting our sales team to new activity (more on that later).
Census delivers the behavioral data we have in Redshift into Salesforce, giving our sales and marketing teams a clear view of our user interaction. Although there are a lot of moving parts, there are clear advantages to building a stack that can deliver insights into our user behavior in both our data warehouse and our CRM.
Delivering richer insights with Census
Product analytics can be full of red herrings. For example, if a user spins up a Snowplow pipeline but quickly abandons it without taking any other action, we could hardly consider them a highly engaged user. Conversely, another user might never deploy a pipeline, yet spends a lot of time browsing documentation, asking questions in our Discourse forum and reading articles on our website, that user might be inferred as a highly engaged prospect.
These nuances are only perceptible when we have access to rich behavioral data. Thankfully with Snowplow, we can capture all sorts of information about how trial users interact with our channels, mapping out their journey in detail from the buttons they click to the time they spend reading documentation. This is great for analytics, but what about helping our sales team make informed decisions?
As we mentioned in Chapter 3, operational analytics is not just about capturing granular data, it’s about actionable insights that front-line teams can use to drive success. With Census, we’re able to deliver this data into Salesforce, which allows our sales team to build meaningful relationships with prospects. By seeing how trial users are interacting with Try Snowplow (for example, by determining which use cases they’re most interested in), our sales team can generate richer, more engaging conversations with them. This visibility enables us to provide better support to prospects, and gain a deeper understanding of the bottlenecks in our product experience.
Having Census relay insights directly to sales reps has an immediate impact on their success. It keeps the sales team informed about the prospect’s needs and ambitions, which makes dialogue smoother for both parties. As for the product team, understanding the user journey enables them to identify potential drop-off spots, or stages where people get stuck.
Operational cohesion: better together
Being able to send behavioral data to and from Salesforce has not only helped us improve the trial experience for users, but also to streamline internal workflows. We don’t have to rely on engineers to constantly implement new tracking, or take up an analyst’s time building dashboards. Better still, our sales team can continue working within the platforms they’re most comfortable with – which makes them far more inclined to put this data to good use.
One of the best outcomes is that our sales and marketing teams can work with our modeled, meaningful data in Redshift, which means we can apply our unique logic to the data and set strict criteria to those users we would classify as “a prospect”. By specifying which actions the individual user needs to take before a sales representative reaches out, we ensure that prospects are only contacted when they’re ready.
Census has been instrumental in allowing us to close the loop on the data lifecycle – bringing sales, marketing and product together to learn, iterate and improve. Because the data is bi-directional, traveling both from and to the data warehouse into Salesforce where it can be put into action, there is a constant flow of intelligence for product managers and sales reps to work with.
While we’re still working on making Snowplow to be easier to discover, trial and experience, the flow of behavioral data into our front-line platforms empower us to make incremental gains along this journey.