Test Results: Promoted Listings on Books comparing values from 2% to 15%

[blurryrobot]

Attached to this message are the results from some tests I did on Promoted Listings.  
I wanted to see whether a higher percentage would result in increased sales. 

I created two campaigns by splitting a larger one in half (paging thru in alphabetical order, sending alternate pages to each campaign to avoid bias).  Set one to 15%, the other to 5%.  

After about three weeks, the 5% campaign was getting about 1/2 the impressions and 1/2 the clicks, but was paradoxically outperforming the other in sales.  At that point I reduced both rates, the 15% campaign went to 5%, and the 5% went to the 2% minimum. 

Both campaigns, interesting enough, registered a large SPIKE in impressions when I changed the rate on 4/15, even though the rate change was a drastic reduction, not an increase.  I wonder, are people increasing their rates, seeing a sales spike, then crediting the increase, when they should in fact be crediting it to simply the CHANGE?  Hmmm....

There is certainly an effect on Promoted Impressions when reducing the percentage.    It is harder to say whether that reduction from 15 to 5% cut down on promoted clicks, though the reduction from 5 to 2% did.  But sales appear to be sporadic in either case.  

I used to promote all my books at a fairly high percentage, reasoning that with books listed by ISBN and being rather standardized, it was hard to stand out, and the promotions could help. 

I am forced to conclude that I was wrong.  Promoted Listings, on books at least, may be drastically over-rated. 

I had also hoped that this test would help me establish a relation between clicks and sales.  But no clear relation is evident.  The reason I had wanted to establish this, is to see what pricing to go for, if I was to try the pay per click ads.  But without any clear relation, it's impossible to value a click...

More tests to follow. 

[wastingtime101]

I haven't tested it, but I've been told the simple act of ending a campaign then starting a new one - even if at the same rate - will result in increased PL performance. Throwing that out there for whatever it's worth to anyone.

[my-cottage-books-and-antiques]

@blurryrobot Because the two programs are so different, I don't think using PLS tests will be helpful to determine the value of clicks for PLA (or how much to bid on the clicks). 

As to PLS, I will continue to follow your tests with interest, but do note a couple things: ebay has been very clear that there is NO clearly defined correlation between the PLS rate and views/sales. This is because the rate is only one factor ebay considers when determining placement for PLS. 

SInce ebay made the recent changes to how they calculate PLS, I have basically pulled back. I still use PLS, but my rates are generally lower than my previous rates, and I am often applying PLS only to some items (in the past, I generally put all items in a PLS campaign).

I sell long tail one offs, and while I believe that PLS (NOT PLA) has real benefit for me with my kind of stuff, at present I'm satisfied with the slow dime rather than the fast nickel. That might change, but for now, I'll continue as I have been doing.

[blurryrobot]

So, earlier today I was talking with a friend who occasionally buys on Ebay but doesn't sell, and I shared these test results with him.  Once I was able to explain what I was talking about, my friend stated that he was completely unsurprised!  He said that he regularly skipped over the 'sponsored' listings and went to the next page, comparing it to the 'sponsored posts' on social media which he had also gotten into the habit of ignoring.  

Now, my friend is not a "power user" by any means, just an occasional buyer.  It was very interesting to hear him say this, in light of the results above. 

[mam98031]

PL listings work very well in some categories and not so much in others.  You are great at doing research and that is exactly what you need to do so you can decide what is best for you.

Good luck.

I don't use them at all.

[simply-the-best-for-you]

---

@blurryrobot wrote:

So, earlier today I was talking with a friend who occasionally buys on Ebay but doesn't sell, and I shared these test results with him.  Once I was able to explain what I was talking about, my friend stated that he was completely unsurprised!  He said that he regularly skipped over the 'sponsored' listings and went to the next page, comparing it to the 'sponsored posts' on social media which he had also gotten into the habit of ignoring.  

Now, my friend is not a "power user" by any means, just an occasional buyer.  It was very interesting to hear him say this, in light of the results above. 

---

Very interesting.  The only issue I have with your test results is that they are yours.  They can't be extrapolated across mutliple users or categories.  They are interesting for sure, but there's no broad based reccomendations that can be made. 

I'm in a super saturated category & PL's TRIPLED my sales initially.  Everyone says they never click on Sponsored listings (I've had a number of buyers tell me that when in fact they'd purchased off the sponsored & never noticed).  The VAST majority of my sales are from PL's.  Hard to say how many would have sold anyway, but based on the fact that my sales tripled, I'd have to say not many.  

It's still interesting to read your results, but it doesn't mean they'll be my results.  I reduced my PL %age in half & only lasted 2 days b/c there were no sales.  It's so category specific.  Thanks for sharing though.  I'm always interested in reading how others do with it.  

[blurryrobot]

Yeah, indeed, this may have limited applicability beyond books I think, it would be interesting to see what happened in your category, which unfortunately I don't have enough items in to run a proper test.