I spend a lot of time running customer research and communicating what we’ve learned to stakeholder teams. A frequent challenge I see stakeholders have is their understanding of observations vs. insights.
Many people refer to everything you saw, heard, or deduced from the customer experience research as insight, which isn’t correct. In this blog post I want to explain the difference between observation vs. insight.
Keeping it simple, I would define an observation as:
Something you saw or heard from participants
And an insight as:
Something relevant to your investigation that you have deduced from multiple observations
Let’s use an example to illustrate this: You’re designing an appointment booking web page for an estate agent and heard a stakeholder say:
“We have some interesting insights from our research. Most participants didn’t notice the book now button, and they generally skimmed past the hero image and primary messaging to go and find X.”
These are just things we have seen or heard (observations). Yes, they are helpful, but you could create many different design solutions to this, with no idea which one will help your user. You could suggest:
- Make the book now button bigger
- Change the color of the book now button to a more noticeable color
- Increase the brightness of the hero image
- Use a more diverse selection of people in the hero image
But will any of these solve the real problem?
If you combine those observations with others, you might be able to generate relevant and actionable insights for the team.
Let’s say you also saw or heard that:
- Participants have a lack of trust in estate agents.
- Participants wanted to see evidence of the company’s credibility before booking an appointment.
- Participants wanted to speak to an advisor to ask questions before making a booking.
- The target audience is less tech-savvy than the general population and feels more comfortable booking by phone than online.
Now, we can start to generate some insight from our observations. In this case, some insights may include:
- Our target audience prioritizes credibility over everything else (due to their lack of trust in estate agents). They need to feel confident in the company before considering booking an appointment.
- While online bookings may save time, participants felt other channels, such as a phone call, could help them complete different tasks (asking questions, gauging service levels) and booking an appointment simultaneously.
You’ll notice that those insights give much more actionable direction than just the observations. For instance, it indicates that the page structure and content should be more focused on building trust than booking an appointment and that perhaps booking online should have less emphasis than a contact phone number.
To summarize, insights should generally have:
- A conclusion: data drawn from multiple observations or themes.
- Relevance: your insights should be tailored to your research objectives and project brief.
- Action: non-actionable insights will not benefit you or your stakeholder team. Where possible, ensure insights are linked to an opportunity for improvement.
While observations are:
- Things you have seen or heard from participants and generally don’t point toward a specific solution.
When going through the analysis from your next piece of research, be sure to map out your observations and cluster them. Then you can create summarized insights that capture what you’ve seen and heard in a compelling and precise manner.