Automated Research: Talking to users, on repeat.
Most technology companies know that talking to users frequently is a pivotal habit that increases empathy, de-risks projects, and surfaces innovation opportunities. At Petal, where “listen with curiosity” and “our job is never done” are company values, this is particularly important.
In practice, though, scheduling and executing user interviews is messy, time-consuming, and can feel like a burden for teams that are used to moving fast. Querying the base for eligible customers, emailing back-and-forth to find the best time, screening, and signing NDA (non-disclosure agreements) docs with multiple candidates at the same time can easily eat away a good chunk of a user researcher’s calendar.
The good news is: it does not need to be like this! The best user research is that which energizes and inspires a team, not blocks it from moving fast or feels like a chore.
When kicking off Petal’s research program, it has been our goal from the start to unlock the power of user interviews by getting rid of the recruiting overhead and making it “just happen”. I partnered with Jack Docal, my marketing colleague and automation wizard, to find a better way.
Instead of designing a specific, expensive research study or topic, for this project, our philosophy was simple: “Let’s get the users coming, and no matter what we show them or talk about, we’ll learn something, every week.”
The workflow
We were able to develop a seamless, automated workflow using a mix of tools: Iterate (surveys), Zapier and Iterable (automation), Google Sheets, Zoho (document signing) and Calendly (for scheduling).
Here’s how it all comes together:
Everything starts on our satisfaction survey, sent to members at different stages of their customer lifecycle. On the survey, there’s an opt-in question that asks them if they’d be interested in joining a research session.
Those who say yes are taken to a quick screener designed to filter out technical folks or colleagues from our own industry. If they pass the screener, they’re sent to Zoho, to authenticate their email and to to sign NDA documents. From there they are sent an email with a Calendly link, where they can find the time that best suits their schedule. Calendly checks times against my calendar, to make sure I’m available for a call, and a Zoom call is automatically created in a public calendar that the whole company can watch. Finally, the users get a confirmation email with instructions on how to prepare for the call.
Results
The results were overwhelming. When we turned the workflow on, with a purposefully reduced funnel volume, 10 users poured in, with signed NDAs and confirmed calendar invites.
Closing this same number of interviews with a hands-on recruiting approach would have taken much, much longer. Ultimately, we had to turn the workflow off to avoid getting too many! The power to flip a switch and instantly get a month’s worth of customers willing to talk to you is every researcher’s dream.
Next Steps
This workflow, of course, does not cover all the user research bases. In order to ensure that our teams have the best evidence they can to make user-centered decisions, we need to create more than a recurring cadence. We need diverse methods, so we don’t get stuck with the ones we’re comfortable with. We need to reach a more diverse audience, in order to get a representative sample of our user base. We need to mix quantitative and qualitative evidence, so our findings reveal the “what” as well as the “whys” behind the data. We need to fight the silos that create duplicate findings or hidden knowledge. And we need to be constantly doing research at all levels of abstraction: from interface micro-interactions to rough mockups of new ideas, to our user’s mental models about their finances, and in the strategic, birds-eye-view level of broad behavior and societal change.
Our data, product, customer support, and marketing teams are working tirelessly on initiatives to create a sound, powerful research program: commissioned and independent third-party studies, automated surveys, contextual feedback entry-points in-app, A/B testing frameworks, data democratization, customer support center logs, our own internal empathy gauge, and many others.
We’re just getting started, so stay tuned to hear more about them all.
