When we started building Reel Reviews, we asked a simple question: why do people read reviews? The answer surprised us. They don't read reviews to confirm a place is good�they read reviews to find out if it's bad.
The Problem with Five Stars
The traditional review system is broken. A restaurant with 4.8 stars and 2,000 reviews tells you almost nothing useful. Is it actually good, or just not terrible? Are those reviews from locals who know the area, or tourists who stumbled in?
We noticed something interesting in our research: when people are deciding where to fish, they don't want to know about the one perfect day someone had. They want to know about the bad days. The days when the fish weren't biting, when the weather turned, when the boat ramp was crowded.
"People don't trust five-star reviews anymore. They're too easy to fake, too generic, too positive."
� Dr. Sarah Chen, Consumer Psychology Researcher
The Psychology of Negative Reviews
There's solid research behind why negative reviews are more trustworthy. It's called the negativity bias�our brains are wired to pay more attention to negative information because it often signals danger or risk.
In the context of fishing, this makes perfect sense. A bad fishing trip isn't just disappointing�it can be expensive (fuel, bait, time off work) and dangerous if you're unprepared for conditions.
What We Learned from User Interviews
We interviewed 50 anglers before writing a single line of code. Here's what they told us:
- 87% said they specifically look for negative reviews first
- 73% don't trust reviews that are all 5 stars
- 91% want to know about specific problems (crowding, weather, fish activity)
- 64% have been burned by overly positive reviews in the past
Building for Honesty
This research led us to a counterintuitive design decision: Reel Reviews would only show negative reviews. No star ratings, no "recommend this place" buttons. Just honest, specific complaints about what went wrong.
We built several features to make this work:
1. Structured Negative Feedback
Instead of free-form rants, we guide users through specific categories: fish activity, weather conditions, crowding, facilities, and safety. This makes reviews more useful and easier to compare.
// Example review structure
{
"location": "Auckland Harbour",
"date": "2026-03-10",
"conditions": {
"fishActivity": "low",
"weather": "windy",
"crowding": "moderate"
},
"details": "Winds picked up around noon, made it hard to keep position."
}2. Photo Verification
Every review requires a photo taken at the location. This prevents fake reviews and gives context to the complaints.
3. Time Decay
Older reviews fade in relevance. A complaint from six months ago matters less than one from yesterday, especially for fishing conditions that change seasonally.
Launch and Early Traction
We soft-launched Reel Reviews in February 2026, focusing on the Auckland area. The response was immediate and surprising.
Within the first week:
- 1,200 downloads
- 400 reviews submitted
- 4.9 star App Store rating (the irony isn't lost on us)
- 68% daily active user rate
The most common feedback? "Finally, honest reviews." Users loved that they could trust what they were reading.
Lessons Learned
Building Reel Reviews taught us several valuable lessons about product design:
Constraints Drive Creativity
Limiting ourselves to negative reviews forced us to think harder about what makes a review useful. We couldn't rely on star ratings as a crutch.
Trust is the Product
In a world of fake reviews and paid endorsements, trust is a competitive advantage. Everything we built reinforced that trust.
Community Self-Policing
We were worried about abuse�people leaving fake negative reviews to sabotage competitors. But the fishing community self-polices remarkably well. Experienced anglers can spot fake complaints, and they call them out.
What's Next
We're expanding Reel Reviews to cover more of New Zealand, and we're exploring similar approaches for other activities. The principles we learned�honesty over positivity, specificity over stars, community over algorithms�seem to apply broadly.
If you're building something, consider what your users actually need to know. Sometimes the best feature is the one you leave out.
