Does the Algo hide items for not selling?

[movieman630]

If so, how long is too long without a sale for the algo to bury something? A week? A Month?

I've noticed a weird trend.  I have over 4k listings.  But a seemingly random item will all of the sudden become very popular, even selling multiples of the same item in one day.

Out of 4K+ items I find this anomaly very curious.  Do they show something that sells at the top of the search results to more people in a given category?

Is there some sort of invisible "hot item" momentum push?

Also, does anyone know why impressions would always be around the same amount average?  Why in the spring when I was averaging 300K impressions a day have I gone down to now 200K a day?  I haven't lowered my PL %, in fact I've raised it. 

What changed?

Why do we get multiple orders for the same random thing when thousands of other listings go unviewed?

I have a feeling the algo isn't "smart" enough to know when items (that might be stagnant) would sell if shown to certain buyers and just ignore them because they haven't sold.  If so, this would be a terrible oversight.  Potential buyers should always be considered for "stagnant listings".  Many things won't be hot, many items are very specific.  But to not show them in results at all is counter productive. 

Follow me:  cellphones are hot, and highly competitive.  The Algo filtration of those items will be aggressive.  There's many brands, styles, and options.  It's something billions of people use, everyday.  Same for clothing.  

However, the same algo should not be applied in Antiques & collectibles.  Why?  Because those items are unique, some very specific.  If the same algo popularity contest is applied to those items as it is in common items categories the results will be a LOT of stagnant listings because the algo isn't showing them.  The algo should be forego'd in collectibles.  It should be more organic search functions in those categories.

[vintage-camerastuff]

If I have several of the same item I notice as soon as one is sold another one follows pretty quickly.  Probably due to watchers seeing one sold and not wanting to miss out.   

[dhbookds]

I don't think it's necessarily an algo.......rather retail ups/downs/popular/not popular.  Social media/magazine articles can impact an "item's popularity" over night......and likewise an item can become stagnant after a big rush.  # of impressions can be attributed to the number of people shopping and the number of items listed.......I would guess sellers are increasing listings for Xmas shopping.......but the # of buyer eyes hasn't increased at the same rate...........

[meme6253]

Not to show them is counter productive, You Hit the nail on the head!

[az93]

Nobody knows.

[mybigsale]

@movieman630 

Does the Algo hide items for not selling?

No (in my opinion).

I believe eBay pushes items up. If your item is not being pushed up, it is sinking below the other items (and not because eBay tied a brick to your items).

The algorithm likes item movement and activity as well as new items. I don't believe it is so much the algorithm is noticing your stagnant items, it is just taking notice of new and refreshed items. Your stagnant items drop to the bottom because of the process of raising up items with more recent activity.

It is a consequence of eBay's positive actions (according to their algorithm) - not the algorithm trying to find items to sink (that will happen as other items rise). Of course, sellers who do not advertise will not agree placing items with Ads above their items is positive... positive action is as determined by eBay.

Refreshing and relisting stale items does appear to be something the algorithm notices and acts on. 🙄

[movieman630]

So selling the same item multiple times in a few days or the same day is manipulation of some kind.  It makes sense but it's really weird when it happens.  Especially when you look at the "need" of the item being on par with thousands of other listings of the same genre. 

It's just odd how one random item gets that much action when thousands of others see none at all.

[robbie31415]

I wouldn't say it's codes that x time as passed deprioritize listing.

However there are a multiple reasons it is boosted initially and then as time goes on newer listings get boosted up.

To combat it you have to create reason for the algorithm to continue to boost which there are multiple ways to make the algorithm happy.

[squinkercat59]

I remember years and years ago Country Living magazine did a feature on a woman that collected antique potato mashers (I subscribed, and read the article myself). I managed a large antique mall at the time, and for several weeks around that time, people came in looking for antique potato mashers to the point there were none left in the store (120 dealers). Coincidence?  Or the article just giving people a new idea.  We'll never know for sure. But enjoy the sales!

[reallynicestamps]

If you make a sale, eBay shows all your items slightly higher.

If you make a sale in Ohio, eBay shows your items to Ohioans.

If some one  sells a potato masher, all potato mashers get more Views and eBay shows potato mashers to more people.

And of course-

If you have more pictures ,or offer Free Shipping, or one day shipping, or ship internationally, or have Item Specifics filled, or have the newest listing, or have just relisted a stale item after resting it for a few days.

Or Or Or

Or you have had the only one on eBay for six years  and finally someone is looking for just that thing.

[redlinear]

I consider it a timing thing. 
When you list an item, it's easily seen near the top of the page of search results.
This is eBay, so...it may not be what was searched for, but, it stands a good chance of being somehow related to what was searched.
Other sellers list related items constantly, so you're item being near the top, doesn't last long.
BUT, sometime in the future, it's going to get a 2nd, and maybe a 3rd chance toward the top.
    Consider that not many shoppers know how to search, and generally don't go down half a page, especially since it's not a list of exactly what they searched for, and they don't know how to sort/filter....if they did, they are tired of looking already, and are going to buy whatever is in front of them if it's reasonable, or go elsewhere to shop.
That is what it seems like happens.

Then again, you're dealing with an algorithm or Ai or whatever...that's smart enough to return a search result that IT THINKS you want to see, in less than a second....all while considering Promotions, Time started, Time listed, Seller rating, and, and, and, and.  
It's pretty darn smart enough to forecast sales I'd bet