Guest Blog: ‘How DOJO AI Helped Us Understand Our Users’
Guest Author:
Francisco Esteves, Country Manager Refurbed.com
This isn't a success story about mastering analytics. It's about learning to question what we thought we knew about our users, and how AI helped us see the blind spots in our thinking.
Breaking Down Our Assumptions
We thought we understood our market dynamics perfectly. Our premium products were performing well, we had steady traffic, and our analytics seemed to tell a clear story.
That was until we started diving deeper with DOJO AI's Data Analysis Agent.
The First Reality Check
The wake-up call came when analyzing what we thought was a straightforward pattern in our category performance. Our assumption was simple: longer session duration and more pages per visit meant better user engagement. After all, if people spend more time browsing, they must be more interested, right?
DOJO AI's analysis showed us how wrong we were.
As we dug into the data, we discovered something counterintuitive. Users who spent less time but viewed fewer pages were actually showing higher conversion potential.
The platform revealed that what we interpreted as "high engagement" (long sessions with many page views) was often a sign of confusion - users unable to find what they were looking for, bouncing between categories, and ultimately leaving frustrated.
The real "aha" moment came when we saw that our most effective user journeys were the focused ones. Quick sessions with direct paths to purchase were outperforming the lengthy browsing sessions we'd been so proud of. We had been celebrating the wrong metrics, mistaking confusion for engagement.
What made this revelation particularly powerful was how it challenged our basic assumptions about user behavior. It showed us that sometimes the data we're most proud of might be telling a completely different story than what we imagine.
In analytics, as in life, the obvious answer isn't always the right one.
Uncovering Hidden Patterns
The revelations kept coming.
Our traffic sources told an unexpected story: while social media drove significant volume, it wasn't converting as we assumed.
The most fascinating discovery was about our weekend traffic. We'd always seen lower numbers on weekends and accepted it as normal. But when we dug deeper, we found that weekend visitors, though fewer in number, showed entirely different browsing patterns. They spent more time exploring, compared more products, and often made more considered purchases.
We'd been trying to "fix" something that wasn't broken at all.
Beyond the Numbers
Working with DOJO AI taught us something crucial: data without context is just numbers. We learned to stop celebrating high engagement rates without understanding if they represented successful user journeys or frustrated customers trying to find what they needed.
The platform helped us move beyond basic metrics to understand the story behind user behavior. We started asking better questions and making decisions based on evidence rather than assumptions.
It wasn't just about knowing what was happening anymore - it was about understanding why.
Looking Forward
While we've uncovered many insights, we know this is just the beginning. Every week brings new patterns to understand and new assumptions to challenge.
That's the beauty of working with a true AI-enabled Marketing OS - there's always another layer of understanding to uncover. And yes, we're still discovering new surprises in our data every day.
