Understanding Amazon's A9 Algorithm
Last Friday, Katherine Khoo shared with BrightonSEO around "Understanding Amazon's A9 Algorithm" within the Amazon stream. She shared a stage with Danny McMillan from Seller Sessions and Prabhat Shah from DaytodayeBay.
Understanding Amazon's A9 Algorithm
Amazon's own data says that 70% of customers don't get past the first page of results for a search. 35% of customers click on the first product and the first three results in search results account for 64& of clicks. That makes these positions highly contested and great online real-estate.
Amazon has been described as that "ruthless meritocracy" by Pat Petriello from CPC Strategy, where "New products have the chance to appear on page one of a coveted SERP as long as they are selling at a velocity and a price which warrants that position. However, established brands still have advantage of customer loyalty and familiarity among shoppers, giving them an edge in consideration and conversion assuming all else is equal. Lead brands also typically have the scale and budget to dominate paid placements relative to challenger brands with smaller budgets.”
Starting off #BrightonSEO with a Talks on Amazon Shopping - first talk by @_KKhoo - slides here https://t.co/NSwaZN7ch9 some really good analysis in the deck, what's important - price or ?? pic.twitter.com/wxCiOy42FI— Gerry White SEO geek (@dergal) September 13, 2019
What is A9?
Firstly, let’s personify this algorithm a bit… What is A9? A9 is actually a separate company to Amazon that is dedicated to search. They’ve got a website, a9.com, a team and a great list of clients.
They are a team that just built search algorithms and manage search and advertising technologies that are scalable, highly available, and cross-platform for our parent company, Amazon, and other clients. And, fun fact, is also used by M&S .. so if you were looking for your Gareth Southgate waistcoat last year, you may have used Amazon’s technology inadvertently.
What known factors are in A9?
Secondly, let’s unpack the factors of A9.
Worth starting by saying that, unlike Google, Amazon do not announce amends and do change their algorithm periodically. The following factors are found through anecdotal evidence and are tried and tested. There are primary factors (Direct) and then secondary factors (indirect) - basically things that affect your sales.
So with Amazon - Direct factors on your ranking are your Text relevancy, Price, Availability and Sales Velocity (or whether your product is fast moving).
Indirect factors are clearly then things which influence this. Let’s walk through these...
Direct 1) Text Relevancy
The best way to think about text relevancy is to envision yourself as the shopper. (As you likely are.) When you or I.. or Isla… are searching for a pair of “green scarf”, it’s likely they will type their query in the Amazon search bar rather than filter through the categories section. Keywords are then matched against the search terms marketer’s have entered for their product.
Most shoppers search for products, so it is important to develop optimised product titles and descriptions to stand out on Amazon’s search engine results page and rank against competitors.
According to Amazon, “The number of views for a product detail page can increase significantly by adding just one additional search term – if it’s a relevant and compelling term.”
Important components of a detail page include:
- A concise and unique title, in Amazon.com title style
- 5 bullet points conveying the most important information about the product
- An accurate description of the product, discussing all its major features in detail
- A clear product image of exactly what the customer is purchasing
Direct 2) Price
We all know that price is a very important buying decision for people, but just how important is it? For 41.9% of Amazon Shoppers, price is still King. And we can relate to this right? Sometimes we just want a product, and will happily settle for the cheapest available on Prime.
You can run an experiment - take the first page of results (typically between 55-65 products) with 8-12 sponsored products in there. and you can compare the average price of the top 20 to the average of the whole page.
I picked some arbitrary products and ran this:
- Green Scarf - average price £9.18 and top 20 average was £7.45 c 18% lower
- Pretzels - top 20 were 15% cheaper
- Heeled boot - top 20 were 7% cheaper
- Marc Jacobs - 7% lower..
Think I’m making my point.
Similar studies have been done on larger sample sets and compared the price of the top 3 listings to the top 75 products… Overall, the top 3 listings had 20% lower prices than the overall average of the top 75 products.
What does this mean? It means price matters. And it matters a lot for rankings… Unless you have an extremely differentiated product, if you want to rank high you most likely need to have a significantly below average price.
What if you can’t change price? Good question as typically as a marketeer, price is outside of your control. It is worth running this experiment for your product and seeing what comes back. This shows your business and client that you're thinking about their business outside of your normal capacity.
Direct 3) Availability
We all know the feeling of wanting a product (for me, the next pair of shoes) clicking on it, finding our size (Size 5) and then… Not Available. This is bad for me (as I love shoes) and bad for you as a seller, as here is a buyer willing to purchase your product and I cannot. Amazon, want to avoid this disappointment.
Amazon care about getting a product to a customer - rather than being loyal to you as a seller. The customer is always the priority. So, if a customer is asking for a heeled boot and you cannot provide one, the sale will go to another Seller who has inventory.
As an SEO, it’s unlikely that your client will expect you to start asking questions about their internal re-stocking and ordering processing systems however, this could genuinely help the health and ranking of their product listings.
You may encourage them to look into their internal processes of how they manufacture (if they are a brand owner) or stock their products, if they are a reseller.
If your client is using Amazon FBA or Fulfillment By Amazon, you can set automatic reminders which will tell you when inventory reaches a specific level so that you can restock.
And, if you’re selling across multiple platforms i.e. a website, eBay, to resellers, you can use multi-channel software which will help you sync your inventory levels so that you are never at zero. This is the land I live in most the time, and it’s amazing how many businesses do not fully utilise software to help solve this problem.
Direct 4) Sales Velocity
Sales Velocity, is basically a measure of sales volume over time. Sales can be observed as helping you move up the ranking, but it is hard to quantify... rather like a black hole.
And are Amazon looking for great spikes in sales? No. It seems that what matters is a steady stream of sales over time rather than peaks and troughs. This is useful to know as it shapes how we come to think about events or promotions related to Black Friday, Prime Day and other external marketing campaigns.
And are all sales created equal? No. It seems there is clear priority given to organic sales through Amazon without any promotions. If you think about it, this makes sense as its signalling to Amazon that your product succeeds in a crowded market.
There are number of factors that clearly affect your sales velocity, product availability etc. However, we're just going to touch upon Reviews, Customer Convenience and Page Interactions/Cart Abandonment.
Reviews - what difference to they make? I'm not sure how many of you have read the reviews on Veet for Men however, I would highly recommend taking five minutes to read and laugh out loud. The top review is probably one of the best pieces of user generated content I have ever read - and it made this product famous. Aside from fame on Buzzfeed, what affect to reviews have?
eCom Crew did a study looking at how reviews relate to product ranking. They took the at the averages related to the search term as review quantity and average varies heavily by the search term/product set. We can see a couple of things. First, the top rankings have way more reviews than the bottom rankings - by a factor of three/four... What is more interesting is that the quality of review doesn't vary a huge degree over all the products on page one. So, reviews seem to help you get onto page one.
Next, Prime and FBA - both help make your customers life easier and get your product to them quicker. FBA helps you in terms of conversions (you'll get the BuyBox more) and Prime seems to hep your all important click through rate from page 1 - we can all relate to filtering by prime as we need that present for tomorrow as we've forgotten...
Finally, page interactions and cart abandonments. What does Amazon do when a customer abandons their cart? It seems that they count this on them rather than you as a seller - so this is accounted for rather like a sale (which is interesting and may not be the case forever!).
There are tools out there for pretty much everything so it's worth knowing what you're wanting a tool to do, before you get into looking around. Firstly, I'd recommend going basic with excel based tools such as Sonar, for finding keywords or AMZ Data Studio for ASIN insights.
Finally, you may consider a real life tool or agency to help you!
To close, understanding A9 is a case of understanding the direct and indirect factors which play… How Text, Availability, Price and Sales are vital and influencing your sales velocity through Reviews, Prime/Free Shipping and Cart Abandonment can indirectly boost your rankings.
Now, come Monday morning, for those on Amazon or with clients on Amazon, you’ll have a structure to assess your products or client’s products against, some tools, and questions in your itinerary to ask that may be outside of your usual remit.
You’ll also have learned some bizarre facts about how much green scarfs cost, and that you should buy and submit a review for Veet For Men… and hopefully some inspiration to re-watch Confessions of a Shopaholic.
Thanks for listening and it would be great to get your questions afterwards or just by dropping me a line (firstname.lastname@example.org)