Re: talking to users first - 100%! I am fairly active in my community and talk with local business owners all the time. This is where I got the idea in the first place ;). Agreed though that doing this formally and generating some user stories before any development could help focus MVP.
Negative review is definitely a big problem and I'm not an expert on that topic, at all ;). If such an app were to have a review system it would be imperative to ensure the trust-worthiness of reviews shown to users.
Though, I'm not sure you would need reviews at first, at all. I generally trust most of what I buy from a given shop X... I trust their purchasing and supply chain departments to have done the research for me. I never think twice about buying X electronic from Target, for example, wondering if that night light for my son's room is going to combust. One of the benefits of brick and mortar, I suppose.
Doordash put up menus on a website and started delivering orders by hand.
I guess I was alluding to the needed(?) network effect for such an app to become popular with consumers.
For me to have this app on my phone as a consumer it would have to have 80% of the local shops I frequent listed.
let's say it is 80, can you imagine a way to get 80% of the supply (stores) onto this platform? Or, could you chart a path to get to x% supply after y weeks?
idea 1: fake it, just list stores and items on a website and find out a way to get users to see it without paying for ads
idea 2: sign stores up one by one, using some interview-derived value proposition, and get your friends to help onboarding new stores, challenge them to sign up x% more stores per week
idea 3: post about it on hn :(
I think this all warrants more investigation!
1) fake review/counterfeit problem 2) support mom and pop
Google Shopping is trying to go after the "amazon prime for the rest of the internet", which seems like a good idea.
On 2), One thing to do would be to do what doordash did, just go to 50 mom and pop retail stores in one region and talk to them. see what their problems actually are.
The fake review/counterfeit goods problem is actually pretty hard to solve, but it's interesting because it's not on amazon's top 3 problems list, so you could disintermediate. Ways to solve it that may be businesses: machine learning, limiting manufacture geo's, relying on video product reviews with trusted influencers, tracking PE acquisitions as a proxy for brand deterioration, reading negative reviews first, etc.