Hacker News Reader

Comment by digisth | Hacker News Reader

"Among the challenges these Goliaths face is integrating inventory and payments systems from retailers big and small that have little experience selling stuff outside of their own storefronts."

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.

View on HN

I briefly consulted for a ecommerce search company.

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.

Giving you a CSV with quantity that is "in-stock" could mean you need a new one in a few minutes at worst and few days more commonly. Consider a model like Groupon.com or Fab.com - one is dumping large quantities quickly and the other is moving small quantities of well crafted / unique goods.

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.

You're overestimating the level of technological sophistication on the retailer side. Many of them not only don't have any "retailer APIs", but they have no internal view of their inventory - so if you want to find out whether a given store has a given SKU in given size, the only option is to call the store and have the guy check in the back.

A middleman service with a standard API plus the push from consumers / front-ends might encourage them to build that out. In any case, I wasn't thinking individual meatspace retail stores; I was definitely thinking more "whatever is sold on the web site now" (e.g. Macy's, Asos', Urban Outfitter's online stores - which I'm making the hopefully-not-wild assumption have their own separate inventory.)

Tackling per-physical-store retail inventory with the challenges you mention is definitely a whole other thing.

There is: https://twotap.com/

disclaimer: I am an angel investor in the company.

Very interesting. Thanks.

Pretty much what we're building at Two Tap, a gateway API that can be used to order any product from a supported retailer's inventory and benefitting from realtime product data (inventory, price etc). A huge benefit to being an independent API is that new features become available for all of the supported stores in a matter of days (vs months-years-never if each retailer would have to update their API).

Reply on news.ycombinator.com