The problem we solve
If you're shopping for running shoes, you've probably had this experience: one expert reviewer loves a shoe, another hates it, and 500 user reviews on a retailer site give it 4.3 stars. Whose opinion do you trust? They're all valid — but no single perspective tells the whole story.
SolePicks solves this by aggregating ratings from multiple expert reviewers and user review platforms into a single weighted consensus score. Instead of trusting one voice, you see the collective verdict.
Our methodology
For every shoe in our database, we collect ratings from up to 8 independent sources spanning expert reviewers and user review platforms — RunRepeat, Running Shoes Guru, Road Trail Run, Runnea, Running Lab, REI Users, Running Warehouse, and Zappos Users.
Each rating is normalized to a 0–10 scale (different sources use different scales — 0–5, 0–10, or 0–100). Expert sources receive a higher base weight (1.2 to 1.5×) reflecting their analytical depth. User ratings carry a weight of 1.0 but scale logarithmically with review volume — a source with 500 user reviews carries more weight than one with 5, but doesn't dominate.
The formula:
Each shoe also has a written summary that synthesizes the strengths and weaknesses identified across all sources. These summaries are AI-generated from the underlying review data, then reviewed for accuracy.
Why this approach works
Single reviewers have biases — preferences for certain brands, foot shapes, running styles, or terrain types. A reviewer who hates narrow shoes will rate the Salomon Speedcross lower than someone who loves precision fit. Both opinions are legitimate; neither is the whole truth.
By weighing many independent perspectives — expert and user — we filter out individual bias and surface what the running community actually thinks. The consensus score doesn't tell you whether a specific shoe is right for you; it tells you what the broad market verdict is. Use it as a starting point, then read the detailed pros and cons to find shoes that match your needs.