Amazon keyword research tool
Before search engines existed, the web was a collection of websites linked to each other, with no way of searching using particular keywords. You had to know a website in a specific niche and follow the links, or alternatively, go to web directories and explore by category (sports, entertainment, news, etc...).
Then search engines emerged, and all of a sudden, users were able to
search for specific keywords instead of exploring forever. Basically, the
problem that search engines solved, was saving time. Users knew what they
needed, but they had to explore websites and niches, and could miss a lot
of information. With search engines, the user has already made the
decision to buy, for example, a coffee machine. All they need to do is
coffee machine in the search box, and hundreds of
products will pop up in front of them.
Amazon Search Engine Ranking Algorithm
Amazon search engine allows users to type in keywords and get a list of products sold by sellers, along with reviews to give the product a sense of credibility (or not), price and description to let the user know what problem the product solves. There is only so many products a user can browse in a given moment, so results are split to pages. Users don't have the time to browse all the products shown to them, they usually look at the products that have the most reviews, appear on the top and meet their needs. So how does Amazon decide in what order to rank products in the search results? Amazon created an algorithm that sorts products, mainly by their relevance to the search term.
No one really knows exactly how the algorithm works, with AI algorithms, it's not even sure that the engineers who designed it know how it works, but factors like proximity of the warehouse where the product is to the user's location is one of them. After all, Amazon needs to be efficient and less distance usually means lower logistical cost.
When you create a listing on Amazon, it asks you to write a product title
that doesn't exceed 200 characters. You are also required to enter bullet
points. In the product title and bullet points you can describe physical
properties of the product (it's a red yoga matt, you would probably want
to include the keyword
red). You also want to include
descriptions of what you can do with the product and how it solves your
problem (you can't do yoga without a matt, or it will hurt your knees).
To summarize, Amazon search engine ranking is affected by the following factors:
- Keyword relevance to product
- Number of returned items
- Number of reviews
- Number of past sales
- Number of positive vs. negative reviews
- User location
- Product warehouse location
To win in each of these items, you need to be on top of your game. You need to make sure your product is high quality, that things don't break fast, that the it is durable and lasts for a long time. Not all products are supposed to last for a long time. There are products that you use one time (for example, soap). You want to make sure your listing description matches the actual product, that it does, look and behave exactly as advertised in your listing.
Each of these goals are won in different ways, we are going to focus on the keyword relevance to product since this is what this article is about. Amazon's algorithm assigns some kind of score to each product and keyword, based on relevance. If Amazon see's that a product gets a lot of conversions using specific keywords, it knows that it probably should be ranked higher. Amazon also benefits from promoting higher conversion products to the top, they get to profit more money.
Choosing keywords isn't the easiest decision. You know what problem your
product solves but there are a lot of ways users are searching for a
solution. Users may be searching, 80% of the time for a
yoga matt but the other 20% actually search
yoga matte. Even though it may seem like it's the same
keyword, using both keywords, or bidding during a PPC campaign on these
two exact keywords can bring in more sales than it would if only one
keyword was used.
The process for coming up with keywords starts in your brain. The best thing to do is to just organize a list of keywords that users may be searching for. The goal is to just write down any possible keyword that could be relevant to your product. A lot of the times, there are multiple ways to search for a product using synonyms. You can use different verbs and adjectives to describe the same product and to target different users.
If you sell in a marketplace where there are already established merchants, one way to find already converting / high volume keywords is to look at your competitors' listings and see what common keywords they are using. Looking at the product title and bullet points should reveal a pretty consistent list of keywords that most sellers are targeting.
Look beyond Amazon - As of 2020, Amazon is the biggest e-commerce platform, but other sites probably have lots of info you can use. Platforms such as Google Shopping, Etsy, and EBay can turn out to be great sources to find more listings and keywords. Simply go on these platforms, type in a few keywords and see if there's additional keywords you need to consider when identifying a market.
Facebook marketplace - In 2016, Facebook released their marketplace. An online platform where sellers and buyers meet and exchange goods for dollars. Although you can find new items, most items that are sold are used ones. You can still enter a keyword into it and examine the products that are for sale, it will give you ideas of what keywords sellers are using.
You can explore websites in your niche by running a Google search and find those suppliers or companies that sell a similar product. Websites normally contain content surrounding items because there are no limits to the number of words and bullet points you can write. For example, you can find useful articles about your item on online digital newspapers much like "The New York Times".
More and more companies have developed external websites for their products and place an Amazon link on their site, so you actually buy through Amazon. The reason is that sellers want Amazon to handle fulfillment. Even though you have to pay fees and commissions to Amazon, you save a lot of money and headache by outsourcing the storage, delivery etc. The lesson is clear - you can extract more keywords from these websites.
A good way to get accurate keyword data is to actually run a PPC campaign with the purpose of figuring out which keywords have the highest number of impressions. Impression is a metric that indicates the number of times an ad was displayed on the user's screen, even if no action took place. The disadvantage of this method is you might have to pay for clicks, if you get any.
The bright side is maybe you'll get some sales, then you'll have more reviews, which will increase your conversion rate. Another approach, is to look at ad copies that have been run elsewhere, like in Google. Perform a search on Google, and evaluate landing pages that contain keywords. Landing pages are usually optimized for keywords to help increase their ranking in search engines.
Persona - a fictional character you create to represent a potential customer of your product. This helps you tailor your product to your users. For example, gym equipment can be used by athletes, and people with disabilities. Each one of them is a persona who is looking to achieve different results with the same equipment, it is a scenario where your product, is a solution to two different problems.
Personas help you find more keywords. To target athletes, you'll want to add keywords about building strength & flexibility, To target people with disabilities, mention keywords specific to that group. Coming up with personas is a productive way to think about and identify your target market for your business, and think about how you solve that problem.
The methods that I mentioned above are useful but they require a lot of manual labor. You have to do tons of research and maintain information in Excel files etc. You can use software tools to save time and make your life easier. With software tools, you can automate (or semi-automate) the process of finding new keyword ideas, filtering out bad keywords and doing a keyword research analysis. Most software tools use APIs from Google or Amazon to check for search volumes and estimate sales per keyword. None of the software tools really have access to the internal Amazon database, hence a lot of the times, you see guesses and best estimates.
Sellerium is a software tool that helps
sellers identify keywords used by competitors. The platform analyzes
thousands of listings and gives you the most common ones used by
competitors. It i a very accurate method and there are no guesses. You can
filter by the number of words in a keywords - one word keywords are
usually not very indicative or competitive, it is usually keywords with 2
words and above that indicate a problem a person has or a solution to it.
back pain indicates a problem where
back scratcher describes a solution. Another filter you can
use is keyword frequency, the more frequent a combination of word is, the
more likely for it have higher search volume, buyer demand, and hopefully
high conversion rate.