Using the Power of AI to Democratize Findings

Yielding user insights and actually using these data to inform strategic decision making are two different practices. While many organizations implement user inquiry, fewer have a good system in place to democratize these insights. What’s getting in the way of high impact user research practice, and how can organizations externalize findings in a way that makes this work accessible?

Organizations are becoming more and more aware of the power of user research methodologies. These conversations can truly be a gold mine, and leveraging this information early and often in product development significantly reduces risk, improves product health, and retains customers. But when it comes to accessing and using insights, we hear this over and over again with researchers.

“One of the hardest parts product design is when you’re working on a team and you are the one that was in charge of doing the research, and trying to get everyone on the same page. Essentially, everything I learned, I want to be able to pass it on to their brain so that we can present together.”

Alisa Moiko, Senior Experience Designer, Digitas

The problem is simple–it’s quite difficult to synthesize and organize these findings. What ends up happening is the brain of the researcher acts as a warehouse. To make strategic progress, stakeholders need to access and apply insights that often live inside a researcher’s brain as tacit knowledge.

When a user researcher embarks on inquiry, they learn through the repeated experience of these conversations. This contributes to a type of deep learning that understands through pattern recognition, is gained via observation, and provides an important source of long term competitive advantage in innovation.

Unfortunately, this type of knowledge is notoriously difficult to communicate. It requires a large amount of processing that isn’t often feasible within current structural organizational realities. Progress is slowed when a stakeholder or decision maker needs to consult the researcher to access these insights. This issue is exponentially compounded the more teams draw on the same research resource.

To make insights explicit, that is, to codify and transfer knowledge to make it usable, is a messy process. Existing solutions include heavy-lift distillations of findings. Conversations with our users indicate that this work can take anywhere from 40 hours to a number of weeks for each set of interviews.

“Then there’s the synthesizing portion that would probably happen over, like, two to three weeks because there was a lot of data.”‍

Kwasi Twum-Acheampong, Head of Design, Twilio Data & Growth

How then might researchers better externalize findings so they can be available at key moments?

Many teams already leverage emerging AI tools for analysis, like Marvin, Usertesting, and ReadAI. While these tools are useful, we continue to hear that researchers need help connecting the dots across their data set. You’d never draw conclusions based on one conversation, so why would you synthesize research that way?

“Qualitative research is messy. It can be hard to dig in and get everything you need out of it. So I’m glad to see CoNote streamlining that process, because even the best analysts are going to miss something in that data, because it’s complicated work.”

Margaret Anne Rowe, Research Analyst, NAIS

CoNote is a powerful user research tool. With a product team of user researchers and dedicated data scientists, it goes beyond summarization of individual interviews you find with other tools. Instead of spending hours revisiting and tagging transcripts, I can use automated highlights as a wayfinder through dense transcripts. Bookmark important pieces of information, side-by-side with elegant video playback and clipping functionality.

As a researcher, CoNote is a tool that helps me and my team see the forest through the trees. Not only am I more able to externalize findings, but automated analysis frees up valuable real estate in my brain and in my day to strategize on how to use these findings.