FP30 and Finding - Episode I (FP30+finding - the Phantom Menace?)

As we get closer to FP30's debut on 9/16, some interesting points are coming out about the new format, how it will work with Finding 1.0 and 2.0 (new search experience). There's also a lot off confusion out amongst sellers. We're hearing definitely incorrect information from sellers coming from message boards, pinks and sometimes even different messages from various eBay groups (the powerseller forums seem to be a great place to get bad info). In our webinar last week on eBay changes we were surprised so many of the questions related to FP30+finding.

This the first in a Saga of posts that will dig into various questions out there and also highlight some strategies based on the answers that we think will give sellers an edge going forward into this brave new world.

Why does this matter so much?

You may be asking: "why are sellers are so obsessed with this topic?".  Since eBay started changing the finding experience with the introduction of BestMatch, sellers have lost more and more control over how their items show up.  Buyers seem to no longer be going beyond the first page of results.  The first page of results has 50 listings and sellers are limited to 10 items/page now so the BEST case is 20% coverage.  The worst case is 0.

FP30 introduces a new element to the equation.  eBay has said that it will be scored in search heavily based on 'recent sales'.  More details on that in Episode II, but for now, think of it as the velocity of the listing - the number of items sold in the listing.  If you have two fp30 listings each with say quantity 100.  If the first one sells 10 and the second 0, the first one (higher velocity) will show up higher than the second one (lower velocity).

BestMatch, FP30 and Auctions - how will the interact?
We know that FP30 results will favor the velocity (time won't really matter, but other factors like seller performance (DSRs), price and what-not still are factors which is important to note) and we know that auctions will still favor time ending soonest (TES in eBay-speak).  So what happens when you have a buyer that searches for 'ipod nano' and there are both kinds of results?

The best analogy I've come up with is to envision two sets of cards - one is ordered by one sorting mechanism (let's say by face value highest to lowest regardless of suit) and the second deck is ordered by a different mechanism (let's say by suit and then lowest to highest).  Once the decks are sorted you need 50 cards out of the 104 total from the two decks.  You don't really care about the other 54 cards, it's those 50 that matter.

There are two considerations in doing this:

1. The mix -The first consideration is the ratio of the two decks.  Do you do 25 cards from each deck?  Do you do 40/10 or maybe sometimes you may wan to do 50/0?  The ratio of deck 1 to deck 2 is very to the end result of 50 cards.  If you are betting on one card and you want it to come out of this process, it becomes pretty important. If your deck is on the 50/0 side, you only have a 2/52 (3.8%) chance of not being there or a  96.2% that it will be there - pretty good odds!  The other direction you have 0% chance (not so great odds).
2. The combined sort -Let's say the mix is 25 cards from deck1 and 25 cards from deck2.  How do we combine the two sets of cards? Do we just stack them on top of each other?  Do we alternate deck1/2/1/2/1/2/1/2, etc?  Do we do 10 deck 1, 10 deck 2, etc. Or do we now re-sort them using one of the first two mechanisms or dream up some new meachanism?

The fact FP30 has a different behavior than anything else on eBay today essentially adds that second deck of cards.  Auctions are deck1 and FP30 are deck2.  If the mix of the first page is 25/25 that's a different strategy than if it's 10/40 or 40/10.  Also remember that you only get 10 shots at this so if only 2 of your FP listings show up, you may want the other 8 to be your auction listings.

While many people argue that FP30 makes eBay strategies easier (I'll put all my inventory on FP30 and kick back and sip cocktails by the pool!) , I think it just got much more complicated and requires some deep thought and analysis.  In fact one of our callers on the webinar made the point that if fp30 is flooded with listings (given .35 listing fee, that could definitely happen in most categories), then ironically auctions could be the best way to get exposure.  One way to look at this is you now have really an entirely new marketplace: eBay auction, eBay store, whatever other channels you support and now eBay-FP30.  Every business will have different needs and ways to leverage/not leverage this new marketplace with it's own finding nuances and fee structure.

FP30 and the Mix/Combination sort
We've spent a good bit of time with eBay's finding team on this topic and here's what we know:

• Mix - The mix will largely be determined for each search term by the buyer behavior for that term (demand-driven).  If buyers for that search term, bid 80% of the time and buy 20% of the time you'll get 10 fixed-price listings and 40 auction listings.  If buyers buy 60% of the time, you'll get 30 fixed-price and 20 auction.
  • One aspect we don't know is the timeframe and how this will act over time.  For example, let's say things are even - 25/25 and buyers start buying vs. bidding, when does that adjust?
  • Also eBay doesn't give the seller the detail on this history so we're all flying blind on this.
  • Finally, let's say you have a product (I'll keep picking on ipods) that gets lots of different search terms (yes that's all products) - some of the terms may have a 25/25 split, others 50/0 and still others 0/50.  eBay doesn't tell you what search term resulted in a product sale so even if you had the term data, you don't know how it flows to conversion.  It's starting to feel like Google AdWords, but you don't know the search terms and you don't know the CPC, but you can see conversions. (More on this line of thinking in Episode III)
• Combine sort - Let's say we have 25 wonderful auction listings that are TES ordered and 25 dandy fp30 listings that are velocity ordered - how will eBay display them to the user?  This is an area of much speculation and all we can get clarity on is that Bestmatch will use it's scoring black box algorithm to do that.
  • Today on the site with FP7 you see a variety of different things and it seems to be in flux in a near daily basis.  Sometimes you see chunks - 5 listings that look TES, then 5 that seem to favor DSRs. Other times free-ship seems to trump all else, etc.
  • eBay is also testing a treatment in F2 that has an auction zone above the listings that pulls out some of the top auctions.
  • My best guess here is that it will be a moving target for a good 3-6 months as eBay tests and tries a variety of different algos.
  • If you don't have a balance of listings out there some of these tests (right now eBay appears to be doing something that pushes fp7 to the back of the bus for example), you can find your sales mysteriously going to zero with no change on your part, or popping 20% with no rhyme or reason, and then dropping 1 week later.

What it all means to your strategy?

• First, auctions clearly should continue to be part of your strategy.  For those terms where auctions have 10-25 listings, you'll want to be there.  If you still want a fixed-price world, slap a BIN on those puppies at the bid price and you'll still get the benefits of auction exposure which is here to stay.
• Second, this is going to require a lot of testing.  If you run some FP30s to see the conversion rate, make sure you run some auctions and you can compare.  Don't assume that you had a 50% conversion rate on auction listings and that will be the same on FP30.  Who knows - it could be higher or lower, etc.
• Third, Given that we are heading splat into Q4, everyone needs to find out what's best for your biz in the next 30-60 days to lock in a strategy.
• Fourth, eBay is sure to change/tweak and retweak things in the coming weeks so prepare to be nimble!
• Lastly, think about the implications of even distribution (10/seller/serp) and de-dupe on your business.  eBay just announced no de-dupe for auction-style listings, so if you have a fp30 and an auction for top sellers those won't be de-duped (see first bullet).  Conversely it makes no sense to have multiple fp30's of the same product as eBay will de-dupe the heck out of them and you might as well burn .35 (well maybe melt a quarter and a dime?).  However, maybe multiple auctions could be smart to fill up to your fp30

In our next Episode...

Phew that's a lot to soak in for one post so let's stop there. I'm telling you though, this is the tip of the iceburg we've got lots more to cover so digest this and stay tuned for more.

In the next post we'll dig into the gory details around FP30's 'recent sales' measure, how it works, how it works with relists and what you can start doing very soon to get ahead of the game.  Stay tuned.

SeekingAlpha Disclosure: I am long ebay and google.

Comments

[Robert]

Nice posts. Even the advanced sellers on eBay are confused on search right now...I think eBay is confused on search right now.

The big issue for me is seller listing costs.

Low initial bids and profitable S&H fees are out the door...so is the excitement of "winning" something on eBay. Now most items are at a price where sellers will not risk much loss.

If I list 1,000 items right now with starting bids between $11 and $40, my insertion cost is about $570. On Sept 16th, I can list the same number of auctions, with just about any quantity per listing as I want, at any fixed price for $350 (1,000 x $.35).

Given a quantity of 10 per auction, I'm now going to be listing 10x more available items for pennies of what it cost me before. Recall that sellers basically paid more for each quantity, or value of listing.

[Ron M]

In the webinar on Aug 27th, "Understanding eBay's Fee Changes" there was a graphic of a table to show the differenct in costs on Ebay for an auction listing versus a FP30 listing. It was showing, based on the item's success rate, the difference is costs of the two forms in listing. Where can I get a copy of that table?

Ron M

[TekGems]

Do you know if sales velocity weight will be based on quantity sold in that period, or number of sales? For example, if a customer buy 100 widgets in one sale order, will that be the same weight as another buying purchasing just qty one? Would my weight be 101 (total qty sold) or 2 (two sales orders)?

[Fruity]

They're going to make adjustments daily based on how they feel the market needs to flow. It seems like alot of wasted time and effort you're suggesting that doesn't work for small sellers. I think it comes down to whether the cost of selling on ebay (inventory,fees and your time/energy) is worth the profit margins sellers get now. Right now, selling on ebay isn't good at all for a low volume seller these days. I'm not prepared to risk my holiday season on the tinkerings of ebay