They assigned it to another engineer. The new "feature" was deployed.
He forgot to round the random number. So let's just say the home page reported the number of active users with very fine precision.
Unfortunately it was quickly fixed.
Booking.com pioneered a lot of e-commerce dark patterns like “X units left” and “Y people viewed this today”.
The EC pushed Booking.com to change its online sales tactics to be more transparent.
It’s the one that does it for the smiles
I haven't been there for six years now, but one of the problems I worked on at Etsy was getting people to stop using their cart as a bookmarking tool. While I was there I worked on the functionality to fave an item as well as the functionality to add items to a List of favorites. There was another tool called Treasuries for this purpose that we phased out, so there were at least three systems that Etsy built to try to help people keep track of items they like without putting them in their cart. I know when we introduced Lists, several colleagues and I worked on that functionality for a few months before it ever saw the light of day. Even so, users continued to put things in their cart "so they wouldn't forget them". It was a very frustrating result; this was the exact behavior we were hoping that Lists would eliminate.
It wouldn't surprise me if it's true that the item really is in that many carts. I would also agree that it's not useful or accurate information for you, another shopper, since an Etsy user "having it in their cart" is in many cases not a very strong signal that they will purchase the item in question. Inflating the number intentionally would definitely be a dark pattern. If that number is intentionally inflated or is known to be inaccurate or fictional, dark pattern for sure.
Whether the current functionality qualifies as a dark pattern or not is a lot harder to judge. Is it poor design? Yes. For sure. Is it harmful to the user? Again, I think yes.
Is it a dark pattern if you design something poorly unintentionally? Does that term measure intent or impact? In a legal context I would expect the measure to be intent, so in that context this is probably not a dark pattern, but HN is not a court of law, so in this context... probably yes?
Inflating the number intentionally would be criminal fraud, plain and simple.
A dark pattern is making the „Delete account“ button smaller and grey while making the „No, take me back“ button huge and green.
Literally lying with the intent of getting money from somebody is fraud and has nothing to do with dark pattern.
Btw not attacking you personally, just wanted to clarify this misconception.
With the amount of Math.floor(Math.random() * 1000) out there, I doubt any case has proven this to be something you can bring charges against a company for.
My wife does this, not just to Etsy, but on all kinds of e-commerce sites. So I apologize on her behalf.
I suspect that it's related that she's also one of those people who will never bookmark a web page. Instead, she keeps a hundred tabs open. I don't understand it. Some people are just wired that way.
It's obvious why this happens. The shopping cart, as an e-commerce pattern, is a bookmarking tool. You put stuff there, it stays there while you browse the store for other stuff. If you step out for an hour and come back, things you added to the cart are still there. If you close the tab and open it later, the things are still there. If the price changes in between, it's updated. If an item goes out of stock, it's reflected on the cart screen. If it quacks and walks like a bookmarking tool, ...
The way for the e-commerce sites to stop it is obvious: put a time limit on the basket. Purge it if the user leaves it be for more than a couple hours. But for some reason, nobody seems to do that :).
As for the browser bookmarks, I also don't use them much (nor does anybody I know). They map badly to actual use cases. For short-term storage, tabs are perfect (especially when the browser saves them between restarts). For mid-to-long-term storage, you want to save your links where you can find them on any device, and often you also want to store them within the service itself (see also: stars on GitHub - they're not an expression of appreciation for the project, they're just bookmarks).
Project maintainers often don’t think so. I think they should be just bookmarks and not a measure of popularity, better to hide the number of stars. It’d hopefully make some maintainers more humble and helpful. Maybe!
And maybe in some cases you would like to know amount of issues per amount of stars.
With issue count if it's a simpler type of library you would imagine there are quite few issues, but you'd still like to know whether this library is validated by enough folks. If it has plenty of stars and very few issues, this seems to be indicative of a very good option.
Ideally you look at all signals, also download counts from packagist/other managers as well.
I just don't think it would be a valid idea to remove the star count in an attempt to make some folks more humbler. I doubt it would make maintainers more helpful.
Seems like the reasonable thing to do in the above case is to have a default "shopping list", and if someone puts an item into their cart and doesn't buy it within the allotted time, it gets automatically moved to the "shopping list". Make the shopping list highly visible on all the cart-related views. Problem solved?
For places that sell expensive stuff like synths, guitars, tech stuff, I add it so I can see the total amount. Then I sleep on it. Sometimes more than one night. It might get bought along the way, so it's a great deterrent for impulse purchases.
I bookmark things in my cart in both Etsy and Reverb because the bookmarking requires an account, and I'm definitely not making one. But at least that doesn't seem to count in the "x users have this in their cart".