Story Beginning:
Peter mentioned this issue to us while filling out a product‑usage questionnaire. He said that when he used the product recommender to search for a model, the “results page didn’t include a parameter comparison,” so even though several products appeared side by side, he couldn’t tell the differences.
In a later study, another user put it even more directly: “I was ready to give up because I couldn’t see the differences between the products and didn’t know how to choose.”
The Real Confusion: Why this wasn’t just “I didn’t understand”
A recommender is supposed to help users “make decisions.”
But if product differences aren’t presented clearly, users must either leave to check documentation or, like the user above, simply prepare to give up.
This means:
A tool meant to reduce cognitive effort was unintentionally increasing it.
Our Assessment: When reviewing the structure of the recommender results page, we found:
• Parameter comparisons were not presented in one place
• Users unfamiliar with our product lineup could not distinguish models intuitively
What made us realize this wasn’t an isolated case was the sentence:
“I was ready to give up.”
When a tool leads users to “abandon making a decision,” the problem lies in the structure — not the user.
Actions and Changes: How customer voices turned into real improvements
Based on the feedback from these two users, we updated the recommendation results page:
• We added parameter comparisons across different products
These changes came entirely from user frustration — we addressed each point directly based on the feedback.
Conclusion: What changed for us because of this customer?
This made us realize:
A recommender is not “complete” just because it displays results — it carries the responsibility of helping users make decisions.
Since then, “Does this help users decide faster?” has become the core evaluation standard for every iteration, instead of simply “How much information can we show?”
