As we dig deeper into what exactly affects search ranking in Amazon we can’t help but take a look at how Amazon treats their own listings. It has often been speculated that Amazon skirts their own rules when it comes to products they sell themselves.

I never personally could verify this until the last few months. As the category for my main products became more and more saturated, I began to notice some trends. First, my ranking was being overwhelmed, but not by third party sellers. More and more 1P products (as opposed to 3P or third party) were ranking fast. Even brand new products with no reviews and limited inventory.

In fact, I saw one product with less than 20 units (that way I could SEE how many sales were taking place per day) surpass me and many other sellers making far fewer sales than we were. That is when I took notice and recalled the many conversations I had with Amazon expert Will Tjernlund about how Amazon doesn’t abide by its own rules.

Then, a few weeks ago, I heard some grumblings around the interwebs about how products with bestseller badges and Amazon Choice badges were given priority in the rankings. While I could see this wasn’t true in every search query, I noticed it in my own category again. Positions one, two, three and four for the main keyword were all products that had one of these badges.

Now I really needed to do some digging….

Hundreds of Listings and This Is What I Found

I decided to manually dig through listings. So I chose as many categories as I could before I couldn’t stare at the computer any longer (I got through eleven of them). In each category I chose five popular keywords and key phrases, and in each key phrase I looked at the top eight listings on page one for a grand total of 440 listings (but four of them were duplicates because sometimes the algorithm is super wonky so it’s really only 436).

I looked to see which ones of these top results were 1P vs 3P. I also looked to see which ones held a bestseller badge or Amazon Choice badge. I even compared the bestsellers in the top eight to how many were on page one at all. First, here are the categories and key phrases I looked through. This was done at random. I did not cherry pick these and I excluded none:

Kitchen
  • Spatula
  • Can Opener
  • Water Bottle
  • Stainless Steel Tumbler
  • Garlic Press
Baby
  • Swaddle Blanket
  • Nasal Aspirator
  • Baby Sling
  • Baby Monitor
  • Teething Ring
Pet
  • Pet Deshedding Tool
  • Pet Gate
  • Flea Shampoo
  • Cat Scratcher
  • Puppy Pads
Home
  • Shower Caddy
  • Toilet Paper Holder
  • Shave Mirror
  • Hanging Basket
  • Picture Frames
Patio
  • Hose Nozzle
  • Sprinkler
  • Grill Mat
  • Grill Cover
  • Grill Brush
Sports
  • Exercise Bands
  • Fitness Gloves
  • Running Belt
  • Yoga Mat
  • Jump Rope
Toys
  • Plush Toys
  • Art Easel
  • Board Games
  • Blocks
  • Kitchen Playset
Cell
  • VR Headset
  • Android Phone Cases
  • Lightning Cables
  • Power Bank
  • USB Car Adapter
Supplements
  • Weight Loss Supplements
  • Hair Growth Pills
  • Vitamin D3
  • Prenatal Vitamins
  • Garcinia Cambogia
Beauty
  • Vitamin C Serum
  • Dead Sea Mud Mask
  • Callus Remover
  • Nail Polish Dryer
  • Beard Oil
Travel
  • Packing Cubes
  • TSA Locks
  • Neck Pillow
  • Luggage Tags
  • Money Travel Belt

What the Data Told Me

First, I want to express that I totally understand this is a VERY small sample size. 436 listings is an extremely small number on Amazon. I also understand there are a boatload of factors not taken into consideration in this small sample. This is only here to serve as an example of what is possible.

So here’s the data:

Of the 436 listings, 157 were 1P sellers. So, 36% of a random sample of listings taken from the top eight search results were sold by Amazon. Now, 36% may not seem like a big number, but let’s consider this; according to research (http://www.ecommercebytes.com/cab/abn/y15/m10/i15/s02) 83% of products available on the mega marketplace are sold by third party sellers.

So, if Amazon only controls 17% of the products but holds 36% of the top search spots, that’s a pretty clear indication they are showing their own products preferential treatment.

Interestingly, however, they don’t dominate EVERY category. There are some they have a huge presence in, and others they seem totally absent from. Some products are also dominated by 1P listings, while others don’t have any. Here’s the breakdown:

  • Kitchen – 42% were 1P
  • Baby – 44% were 1P
  • Pet – 67% were 1P
  • Home – 51% were 1P
  • Patio – 30% were 1P
  • Sports – 20% were 1P
  • Toys – A whopping 72% were 1P (and this is likely the biggest category, so it makes sense)
  • Cell – 17% were 1P
  • Supplements – 32% were 1P
  • Beauty – Only 2% were 1P
  • Travel – 15% were 1P

But what does this tell us?

Unfortunately, it only creates more questions than answers. However, these questions may be good for your business to explore.

For example, who is easier to compete with in terms of taking market share; Amazon or other third party sellers? If the former, then getting into a category dominated by Amazon may be the key. Since that clearly means not many third party sellers are there. If the latter, then you have many options where Amazon’s presence is barely felt.

Another interesting thought is, if Amazon ranks products it sells, then perhaps developing a vendor relationship with them (if you sell one of the products they actively rank for) is an option to consider. In my category where it is being dominated and overrun by Amazon 1P, it may be wise to consider the possibility of throwing my hat in that ring.

And What About the Bestseller Badge?

What benefit all these “badges” have on a listing has been a point of debate for time immemorial. Some say they are crucial to conversion increases. Others claim they have seen very little impact. Regardless of those individual experiences, it is good to dive into what the data tells us about them.

Out of the 436 listings there were 68 total bestseller badges. There were actually 100 total bestseller badges on page one, but 68% of them were within the top eight. Of those 68, thirty (or 44%) were Amazon 1P listings.

This tells us that there are certainly a lot of bestseller badges within the top search results (but that may be a RESULT of ranking, not a factor). It is also interesting to note that more 3P sellers had bestseller badges. However, they were almost split (44% vs 56%) yet Amazon only sells 17% of the products.

In fact, if we take these percentages and extrapolate them over all of Amazon, this is what it says:

There are close to 500 million products on Amazon. That would mean 85 million are Amazon 1P and 415 million are third party or vendors. That means 115 million listings have bestseller badges on page one, and 78 million are within the top eight of search results. Of those 78 million, this would mean 34 million are Amazon’s. That makes 40% of all of Amazon’s products bestsellers. WOW.

Is the Amazon Choice Badge the Same?

The Amazon Choice data paints a different picture. For those of you unaware, Amazon Choice is a tag that makes shopping via voice on the Echo a simpler process. By that logic, based on sheer numbers, you would think a lot of Amazon’s 1P bestsellers would end up with a Choice badge. However, according to our small sample, of the 22 Choice badges only six were Amazon 1P listings.

That means 72% of the Amazon Choice badges were held by third party sellers. We are still speculating on the implications of this, but it is a promising statistic. Although only 5% of the top listings had those badges, so it doesn’t appear the privilege affords a rank boost.

 

What we’ve learned here is not really anything new; Amazon is a powerhouse that throws their weight around in its own playground. However, it is still interesting to uncover the parts of Amazon where they don’t really venture too much. And knowing where they do play lets us as brand owners adjust our strategies.

And remember, when it comes to bestsellers, only Amazon creates more of them than SixLeaf does. 🙂