Regards,
Please, report 
EBay's own category statistics imply there's an issue even if you do analysis of a single variable.
Note, this is not OUR statistics, this is the CATEGORY statistics.
Now I would be willing to bet anyone here that these statistics are for organic CTR only, and that the CTR of PL have increased?
As evidence, if I compare the CTR of our Promotions from June 29 2021-July 29, to our current promotion CTR, it has INCREASED! I compare the CTR of our overall traffic from the same date, it has DECREASED by a huge percentage!
So... This isn't "unusual" and doesn't imply that the search is less accurate in targeting what the sellers are searching for... UNLESS it's a PL?
@babclassics wrote:
velvet@ebay wrote:Hello all! I’ve heard back from the team investigating the concerns about a drop in sales and they said the following:
The change to view counts has done nothing to impact raw view count numbers. Listings are still going to get the same number of views as they would have previously, subject to site-wide trends. We have not impacted or changed anything on the buyer's side. The change in how we count views itself doesn’t impact buy-side impressions, page views, or conversions. It’s just a selling reporting tool that filters out what we consider to be inhuman traffic, and reflects that number to our sellers.
However, we recognize sales as a whole have decreased a little. We speculate this is related to the end of lock-downs across most of the world as we see this across the board with online sales.
Where do we go from here? We’ll continue to run site-wide analyses from a Data Analytics perspective. If anything outside of the normal and somewhat expected decrease in sales given what’s been noted above, then we’ll work to implement measures to solve for whatever is found, in an effort to ensure we’re doing what we can to help connect buyers and sellers.
Yes, thank you for getting a response on this, really appreciate it. But there needs to be (at least 1) follow up. I read and re-read the statement bolded above and I do not see an answer the one of the most basic questions being asked.
We want to know if "non-human traffic" (bots) are being allowed to view and index the entire page and links on that page when they don't count as a view? The answer given is not clear. I agree that the view counts are showing less traffic, but is that traffic still allowed to complete its view of the page? Or are those bots being kicked out of the page once they are determined to be a bot? The answer from IT could be read either way!! They kind of dance around it by saying "buy-side impressions" are not effected. Is a bot considered "buy-side".
The bots need to be allowed full access even if they don't count as a view, so are we 100% sure that is happening? Could it have been programmed incorrectly causing early exit from page being rendered? Exactly how is the view count being decremented? Having been in IT for 20 years i can tell you the easiest way to not count the view would be to just STOP rendering the page output and simply exit the page before the count gets incremented. This would reduce the view count but also stop the bot from seeing the whole page and / or all the links on that page which bots follow. Please ask a follow-up question. Thanks much.
The changes in statistics also need to be considered by the algorithms. As the measurement of statistics was already a mess, but the resulting numbers will need to be updated.
For example, if an item was getting 1000 views (200 of them bots). Now it's getting 800 views. How is the conversion rate being measured? Because if it's 800 views and the same number of conversions, it's going to look like the conversion rate has increased when it really hasn't. This makes your question all the more important.
But if the conversion rate is still counting the original view count, which view number is being used by the algorithm? The one with views, or without views?
I can go to the traffic page right now and find different numbers of views all over the place. This alone indicates there IS major issues, and that a wrong variable is placed somewhere. So why hasn't it been fixed despite being reported awhile back now...?
The traffic pages have matching off-eBay numbers, and matching promotion numbers... but give us THREE different Organic numbers.
In no way you manipulate the numbers being given, do they add up correctly to what is being stated in the daily totals, OR the graph.
And if we download the Active Listings Traffic Report, it gives MORE different numbers for Organic views.
So what is the real number of organic views for the day? Is it 1302? Or 998? Or 778? Or is the FOURTH number in the Active Listings Traffic Report the accurate one?
More importantly.... Which of these is being used to calculate CTR? Which of these numbers is being used by the algorithms? The variables are obviously mixed up somewhere.
But there's no problem, right? I guess we're SUPPOSED to be given correct numbers for Promoted and off-eBay views... but we're NOT supposed to get correct Organic numbers.... And we're supposed to believe this doesn't indicate any problem? It's just "lockdowns across the world"...
Edit: Here's a shot with mouse over for more clarity:
@siamjane8 wrote:
Who can you reach out to at ebay that will take this problem seriously? The response you received back from the team you reported to obviously thinks everything is ok. They are wrong. period. They are not capable of seeing the issue and that is fine- they are not the right team to be corresponding with then.
Somebody or some team is at fault for this issue- if the problem is that they don't recognize or realize what their actions have done- then its time to get new people involved.
Who has oversight over all the IT? perhaps that person can realize what has happened. Who took Harry Tempkin's position?- wasn't he heavily involved in all these changes before he left or got fired?
This is no doubt a multifaceted and complicated problem but i refuse to believe that it cannot be addressed- somebody broke ebay and somebody needs to fix it.
Hi @siamjane8! I can assure you that specific accounts compiled from here and other threads were reviewed by the appropriate team. That research prompted the response they provided me, that I then shared above.
