As innocuous as the term “honeymoon period” sounds, it has been the topic of much contention and debate. Interestingly, most people agree on the functionality of Amazon’s algorithm that gives rise to this phenomenon.
What they disagree on is simply calling it a honeymoon period.
But I digress. Onto the reason you’re reading this essay; what IS this thing I keep referring to as the honeymoon period and what does it mean for Amazon sellers?
Learn all about Amazon's A9 algorithm, the "honeymoon period," and how to leverage these things for greater keyword rank in our comprehensive 70 page white paper. Just subscribe to download NOW! ⬇️⬇️⬇️
HISTORY AND PRINCIPLES OF AMAZON'S A9 ALGORITHM
I remember it vividly. In 2015 there were rumors and speculation, often worded as “do you think Amazon shows favoritism to newer listings?”
Nothing had been tested yet or well documented. And honestly at the time I was unconcerned. It didn’t seem likely and there were more important things to focus on.
Then, six months into my own selling journey, suddenly my well established rank had dropped (position 6 to position 11).
This caused me to look deeper…
I tracked keyword movement for my own products, my colleagues and clients for a period of a couple of months. I kept seeing a consistent trend; after about six months, rank for well established keywords dropped…?!
Working for a launch service I was able to see keyword rank movement after launch, after maintenance, etc. and this allowed me to understand how much volume was necessary for ranking various keywords in different categories. This is how I was able to really dive into the changes after this six month period.
So as I looked deeper at the seeming six-month phenomenon I started making a few other observations. For example it appeared more volume seemed necessary to maintain rank in several categories, and high rank was much more difficult to re-obtain if it were lost.
That’s when I had a poignant conversation with a colleague. In that conversation, it hit me like a lightning bolt…
Amazon uses moving averages as a ranking factor.
What does that mean?
We knew Amazon factored hourly sales (seeing rank increase after only one solid hour of massive sales volume tipped us off to that).
We knew they factored daily sales (rank increase also maintained after a strong week which indicated daily performance was a factor).
Now it appeared that they also factored a six month benchmark.
THAT MEANT that a listing without six months of sales didn’t have a number to plug into that average. So it was a null value.
As in…it didn’t count for or against.
Most new listings have somewhat of a rocky start. Usually that first six months has its ups and downs. Ultimately the average is somewhere in the middle.
Which, if compared to the massive traction offered by a “launch” (or the results of one), can seem low.
So the original “honeymoon period” wasn’t Amazon showing favoritism, but rather the absence of a poor six month average bringing the overall “number” down.
I shared this discovery, and it seemed to make sense, because it was spread and referenced from then on in most Amazon circles.
The six month period, as well as other benchmarks, don’t appear to be as hard and fast anymore. Likely these timeframes have changed, and some speculate anywhere from 30 to 90 days being the new benchmark with the highest “rank weight.”
We haven’t tested for this, but suffice it to say that initial sales velocity still appears to benefit new listings.
The reason we haven’t tested this (in a long time) is because of the emergence of a more recent SECOND honeymoon period.
Yes, with regard to advantages a newer listing may have, insofar as it can rank for keywords with less units and less effort, there is a new “honeymoon” player in town.
Once upon an exhausting Monday evening (I lived in Taiwan at the time, so all my interactions with the US took place from around 11pm to 2am) I contacted seller support to ask about “forbidden keywords.”
I was curious, because I kept seeing people in the FBA-verse (is that a thing? Cuz if not I am making it a thing) get their listings suppressed and in many cases it was revealed they had a single word or set of words that were prohibited in the listing.
While seller support wasn’t super helpful with regard to identifying a list of prohibited keywords (cuz, that would just be too helpful) I did learn that the browse node determines what keywords a listing can be indexed or rank for.
See, some browse nodes don’t allow certain keywords at all. The browse node (or category/subcategory), which is informed by the item-type-keyword, determines keyword relevance.
So the feed-product-type and item-type-keyword which categorize your product also have a set “list” of keywords that Amazon considers relevant for products within them.
This makes sense. It keeps keywords like “shoe” from being relevant to a product in the grocery category. It can, however, have issues.
For example, when I was selling a can opener with a bottle opener attachment, I discovered that so long as my product was in the subcategory “can opener” it could not rank for the term “bottle opener.”
Yes, my listing indexed for it. It was in the title. But Amazon considered it completely irrelevant and therefore any rankability was non-existent for that term (or related terms). This is because of the structure and limitations of categories and their relationship with relevance.
After much testing, observation and collecting data we’ve seen without a doubt that Amazon tweaked their ranking algorithm to be essentially controlled by keyword relevance.
Every factor considered in ranking ultimately determines how relevant a product listing is for a keyword, and thus how rankable the listing is for that keyword.
So, rather than things like conversion rate, sell through rate, time on page, listing activity, etc all directly impacting rank, they now actually impact relevance, which impacts rank.
The second “honeymoon period” is the time in the beginning of a listing (or rather, the beginning of text input into a listing that can be crawled by Amazon’s “spiders” and thus indexed) where without relevance history, relevance is automatically granted.
Also, almost all keywords input in the listing are indexed. So almost every keyword entered into a listing during this time will be indexed and given the benefit of the relevance doubt, making it much easier to rank for those keywords.
This period only lasts a couple of weeks at most, as listing activity, traffic and sales will help Amazon determine a relevance score for each keyword and phrase. But in the beginning, that listing activity, traffic and sales can be purposefully guided to help in this process.
Hence, the second honeymoon. A “Relevance Honeymoon.”
As if it wasn’t obvious, the way you utilize this is to make sure your marketing is aggressive early on. This means starting PPC on auto campaigns with high budgets so you force Amazon to search and find relevant keywords for suggestions.
It means running promotions and rebates from multiple traffic sources to continue telling Amazon that all the keywords you are targeting are relevant. It means building up as much activity, traffic and sales as possible as quickly as you can so that your listing remains highly relevant and primed for rankability across all key phrases.
So you should map out your entire marketing plan before going live. Map out promotion strategies (and number of units), PPC campaigns, influencer campaigns, etc. Make your launch a well thought-out strategy rather than a launch-and-pray technique.
Then, enact that strategy as soon as you have inventory available on Amazon.
When does the honeymoon period start?
In our experience it starts the moment product is available for purchase from the listing.
However, it isn’t clear whether this is true for both types of honeymoon period.
So it is best to keep your listing as bare bones as possible (or even avoid creating the listing) until as close to “go live” as you can.
Can you reset honeymoon?
Yes, but some methods are easier than others.
Relevance reset is easy. Just take all of the words out of your listing (except for the required ones…at least one word in the title, a letter or character in the first bullet, etc).
Save the listing, wait about 20 minutes and check to ensure you aren’t indexed for keywords anymore (anything that used to be in your listing that isn’t). Then refill it all back in!
This forces all keywords to be re-indexed as NEW, and the listing must go through the process of proving relevance again.
Resetting sales history is a bit more complex. Many techniques are taught.
Open a new variation and then merge.
Create a new ASIN and then transfer inventory.
I don’t use (or teach) these methods, but they do exist and shouldn’t be hard to find if necessary.