There are two questions you should be asking yourself:

1. Will a person in your target market, when cherry-picking a couple examples they care about, say "Wow, these results are way better than what I would've found myself!"?

2. Are there enough people in your target market to justify the economics of your system?

Answer #2 first, and then put all your effort into making #1 come true.

Google worked because everybody searches the web, and Google was clearly better than rival search engines. Many metric-based businesses fail because not enough people care about the metrics to make enough money; the remainder fail because they're not a noticeable-enough improvement on the status quo.