Auctions experiment, a big FAIL

[ricarmic]

I thought it worthwhile to share this one to see if other's experience is similar or different.

Experiment was to "re-run" auctions because I noticed auctions were generally getting more views than BINs in the hopes it brought more eyes to my auctions and BIN material.

My normal approach is run item at auction for 7 days then move to BIN. Note that starting price for the auction is the same as would be the case when it is moved to the BIN.

So the experiment was to run each unsold item repeatedly for a month, so repeat 3 more for unsolds. Each time I used relist. After the month, as usual gets listed at same price as BIN via sell similar.

Normally auctions sell in the 10 to 20% range. At my best the "normal way" I've got 50 to 70 auctions running.

During the experiment was typically running 200ish concurrently. Sell rate was down in the 2% to 3% range. 

I've long felt that auctions don't cause sales when they're being put in. In far too times to be coincidence stuff sells whilst I'm listing BINs.

While relisting 20 or 30 items a day over several weeks nary a sale while relisting auctions. (Still sales happening whilst doing the BIN listing which is a sell similar). so I think this confirms at least in my stamp world the algorithm doesn't "count" auctions in the favour the seller list, originally or when relisted.

The impressions counts of the auction reruns were abysmal, 0 or 1 or 2 predominantly.

So I put the exhorbitant recommended standard promoted level on them for the last week. So far not one sale on a promoted auction. View counts are higher, sometimes higher than the first week but no sale.

As the quantity of rerun auctions grew, with generally low views, it seems that the new, first week auctions are also "depressed" in terms of views too. So my theory is that the bots see my auctions aren't popular and are lowering my visibility on them all and perhaps the store too.

I do see that bids definitely attract more views and bids. I had rerun items with zero bids sell signficantly above start price the 2nd or later run when there was an early bid. Still not enough to provoke me to retry a 99c start experiment.

So this has been one of my worst experiments, short summary of my beliefs (which may or not be true):

  Listing auctions doesn't foster more sales during listing (whether original or relisted)

  Listing BINs does foster more sales during listing

  Rerunning auctions using relist doesn't spur more views/sales

  Bids on an item does cause increased views

  Running too many auctions with very low view counts may depress views for auctions and BINs

The results were abysmal enough that the question as to whether relist or sell similar would make a difference I probably won't try until I've exhaused other future experiments.  

Interested to know how these beliefs align with other's experiments, experiences.

[lotzofuniquegoodies]

---

@ricarmic wrote:

I thought it worthwhile to share this one to see if other's experience is similar or different.

Experiment was to "re-run" auctions because I noticed auctions were generally getting more views than BINs in the hopes it brought more eyes to my auctions and BIN material.

My normal approach is run item at auction for 7 days then move to BIN. Note that starting price for the auction is the same as would be the case when it is moved to the BIN.

So the experiment was to run each unsold item repeatedly for a month, so repeat 3 more for unsolds. Each time I used relist. After the month, as usual gets listed at same price as BIN via sell similar.

Normally auctions sell in the 10 to 20% range. At my best the "normal way" I've got 50 to 70 auctions running.

During the experiment was typically running 200ish concurrently. Sell rate was down in the 2% to 3% range. 

I've long felt that auctions don't cause sales when they're being put in. In far too times to be coincidence stuff sells whilst I'm listing BINs.

While relisting 20 or 30 items a day over several weeks nary a sale while relisting auctions. (Still sales happening whilst doing the BIN listing which is a sell similar). so I think this confirms at least in my stamp world the algorithm doesn't "count" auctions in the favour the seller list, originally or when relisted.

The impressions counts of the auction reruns were abysmal, 0 or 1 or 2 predominantly.

So I put the exhorbitant recommended standard promoted level on them for the last week. So far not one sale on a promoted auction. View counts are higher, sometimes higher than the first week but no sale.

As the quantity of rerun auctions grew, with generally low views, it seems that the new, first week auctions are also "depressed" in terms of views too. So my theory is that the bots see my auctions aren't popular and are lowering my visibility on them all and perhaps the store too.

I do see that bids definitely attract more views and bids. I had rerun items with zero bids sell signficantly above start price the 2nd or later run when there was an early bid. Still not enough to provoke me to retry a 99c start experiment.

So this has been one of my worst experiments, short summary of my beliefs (which may or not be true):

  Listing auctions doesn't foster more sales during listing (whether original or relisted)

  Listing BINs does foster more sales during listing

  Rerunning auctions using relist doesn't spur more views/sales

  Bids on an item does cause increased views

  Running too many auctions with very low view counts may depress views for auctions and BINs

The results were abysmal enough that the question as to whether relist or sell similar would make a difference I probably won't try until I've exhaused other future experiments.  

Interested to know how these beliefs align with other's experiments, experiences.

---

@ricarmic

I haven't run an auction in ages. The last 2 or 3 I did run that sold, sold at the opening bid. Either the penny or 99 cent test variety. In doing some recent research on a foreign coin I was planning on listing found 1 that sold with 7 bids. Opened at a penny. Sold at 17 cents. FWIW I believe that was US funds. Betting the fees outweighed the profits. Buyer was thrilled. Seller not so much. When I have considered running an auction starting with a reasonably high opening bid was a non starter along with including a reserve. I see majority of your auctions either sold with 1 bid or max of 6 or 7. Looks like you did very Chinese Emporor lot. Congrats!!! And as mentioned previous it has been ages so not familiar with how the fees end up on auctions vs Buy it now. To my mind anything that a seller considers now for auction has to be either super rare or high in demand. 

Did you play around much with end times? That was the other challenge unless you thought it was worth scheduling. When I was testing I was finding it was very easy to get nickel and dimed with surcharges so gave it up and went all in on BIN.

-Lotz

[ricarmic]

I have not noticed end times having a huge factor either way anymore. In the good old days I think they mattered a lot.

When I start auctions today I usually start them to end in evening if on .CA and earlier on .COM (depending on how many international bidders are likely).

When ebay made scheduling free I presumed that meant there wasn't much advantage anymore.....

[msau4301]

I have always run auctions (with very few Buy it Now) and that has always been what works best for me from day one until now (11 years).

I think it varies hugely on what it is that you are selling, and what others selling the same are doing.

In my case I sell mostly vintage Barbie's and the auction route has been fairly successful for me. My competition is a mix of auctions and Buy it Now but I know as a collector myself the auctions can be quite satisfying when you win of course. I usually run about 20 items at a time and chose 5 day auctions.

I've had more sales activity lately, even though I do everything pretty much the exact same, it is dependent in my case on the particular items I'm listing. I always try to have at least one item I know will be popular to get people looking. In the past 30 days I've had 39 sales which I think is fair given I only list 15 - 20 items at a time (per week). Currently I have 17 auctions going, 5 of them have bids, all have watchers and I find the most activity happens just before they end with the last minute blasts (many are ending tomorrow so I'll see if my luck is continuing!)

When I have tried Buy it Now they get a bit of views when first listed then completely die off. I've never used/paid for promoted listings ever so I don't know if that would make a difference but as long as I'm still selling stuff I haven't worried about trying it. If any of my items don't sell after auction I reduce the price slightly or lump them together in larger lots (again with auctions).  There have been many times I've lowered the price, and the auction ended up going higher than previously listed.

I think everyone has to do what works for them, stamps are completely different than vintage dolls but they are both "collectibles", if someone is looking for something specific we have I think they'll find it either way, the harder part is finding those people who didn't know they wanted what we had until they saw it.

For now I'm continuing as I have with my 5 day auctions (in spite of reading many times here others think it is ineffective), for me it is the only things that works. Plus the added bonus is this 5 days cycle allows me to take random weeks off if needed for holidays, time for family stuff,  etc.

And hopefully my followers watch out for my return (cause I'm taking time off again soon!)

[ricarmic]

I probably should clarify that auctions will remain a very viable portion of my "bag of tricks".

Running the auctions one round only has generally worked well, and it also makes sure I didn't underprice anything (doesn't protect me from the opposite though).

Re-running them en-masse piled up the "slow" stuff which I think also affected the "good stuff". I too have "guaranteed sales" and some of those did not sell on the first round during the experiment which is generally very rare. That's when I began to suspect the experiment failure was causing greater issues than I expected.

[msau4301]

100% I agree that once there are any bids at all on any listing(s) this has a positive affect on your other items being brought forward on search for sure (I've seen this time and time again myself). So I guess on the other side of the scale if you have a large group of re-runs as you say, without bids, it might have the opposite affect.

I can't possibly relate/compare to your very large listing scale, you have thousands (wow, respect!) of items  but if you are noticing a trend on your experiments take heed, no one knows your business better than you!

[ricarmic]

As a postscript, I'm back down to running auctions one round only, have been sell similaring the lump of non sold multies 10 or 20  a day, and sales are going back to the way they were before. 

Today's been unsually strong for a Thursday.

Auctions selling vs listed is back up to 10% or more at the moment again.

So it does really seem that rerunning loads of "unpopular" auctions was dampening more than just the auctions themselves.