Thousands of Abandoned Shopping Carts is a Bot

[engine]

A WSJ investigation into thousands of abandoned shopping carts by a mystery shopper, John Smith, turns out to be a bot, and that bot comes from Google.

The purpose: making sure the all-in price for the product, including tax and shipping, matches the listing on its Google Shopping platform or in advertisements. It wasn’t to cause angst to merchants due to thousands of abandoned carts.

“We use automated systems to ensure consumers are getting accurate pricing information from our merchants,” a company spokesman said. “This sometimes leads to merchants seeing abandoned carts as a result of our system testing whether the price displayed matches the price at checkout.”

[wsj.com...]

[lucy24]

When shoppers abandon carts, websites typically send an automated email prodding them to finish the purchase.

Word to the wise: If I get a nagging email, it decreases the chances that I will come back and complete the purchase. (Major Online Retailer lets you keep things forever, and it doesn’t seem to have done their bottom line any harm.)

He has refrained from blocking the internet protocol address associated with John Smith, of which there are also many

Well, don’t leave us in suspense, wsj. Do those IP addresses belong to Google or not? (And if yes, how on earth did it take anyone more than five minutes to realize it’s a Google-related activity?)

[engine]

It is Google, and Google confirmed it.

[not2easy]

Google owns a huge number of IPs - many of them host unwanted traffic. There is a narrow range of IPs for their "normal" Googlebot. I think it would be very helpful to know what IPs their scripted cartbots come from. There are CDNs that regularly block large ranges owned by Google. Not every site is using rdns or header logging to see who's who. The question isn't who did it, it is 'from where?' in this case.

[lucy24]

I can't help thinking that if the story had appeared in a tech-oriented publication, they would have said something more specific about the IP addresses. It isn’t just Crawl Range vs. Other. It’s also “IPs Used for Recognized Googloid Functions” (GSC-related stuff, faviconbot and so on) vs. “We’d really like to know more” vs. “Sure, they own this range but they won’t take responsibility for anything it does”.

I hope someone hereabout with an ecommerce site can shed further light, preferably from raw logs rather than analytics. Come to think of it, why on earth would this kind of robot bother requesting analytics files at all, as implied by the article?

[engine]

Yes, it would be good to get the IPs.What we need is someone that has these John Smith sign ups to help us with that information.