It stunned me how many online stores selling millions of dollars of goods per month online are essentially incapable of giving you a csv with product id, quantity in stock, and price.
Somewhere in the shit-mass of code that is their site there must be some way to determine precisely what items are in stock and what the retailer would charge for them right now, but it's apparently incredibly difficult to export that : rolleyes :
Our solution was a combination of a semi-busted daily/hourly catalog export, a site scraper, some machine learning, and some human intervention. It was an enormous pain.
ps -- there are definitely ecommerce sites out there where the price you pay depends on your navigation path through the store. I remain unsure if this was intentional or accidental.
The best attack vector is obviously approaching people like prestashop, 3dcart, tictail, etc and trying to get them all to agree on a format to put pressure on shopify. This would enable pinterest to get tens of thousands of merchants on their site with little friction or data rot.
Tackling per-physical-store retail inventory with the challenges you mention is definitely a whole other thing.
disclaimer: I am an angel investor in the company.
Seems like there could be an opportunity for a service to act as a broker; they could act as a go-between for inventory related tasks and services between the front-ends (Pinterest, FB) and the backends (retailer APIs.) Provide a standard API on both ends (so retailers can update prices, number in stock, tax and such, and front-ends can push buys and get updates.) If they were front-end agnostic, they could save the retailers a bunch of time, and handle API updates all in one place (preventing mass breakage when APIs change, since a good part of their job will be to keep the interfaces up-to-date.)
Retail API broker or something. I wonder if anyone is thinking of doing this already.