In this two part niche keyword research tutorial we will be focusing on how to do proper keyword research for an Amazon affiliate niche website. This is part of an overall series which is documenting how to create a successful Amazon affiliate website using a very specific strategy. In the last post we discussed how to find a killer niche site idea by using the Amazon best sellers listings.
To recap, at this point we have:
- Found a product niche using Amazons best seller categories. We were able to come up with water filters as our niche website example.
- Verified that there are at least 90 products on Amazon that could be reviewed in 30 articles. We will actually need about 74 total articles to cover a years worth of posts, but only 30 of those for the first month. This is following our Amazon affiliate niche site strategy.
- Double checked our sanity to ensure that we can actually write that much quality content in one niche without going crazy.
Importance of Proper Niche Keyword Research
Here is where the real work begins and where most people fail from the starting gate when trying to create a niche site. I cannot emphasize enough how important doing proper keyword research is.
We need to ensure that this seemingly golden goose of a niche has not already been beaten to death by other like minded affiliate marketing entrepreneur’s. Though, if you do your research correctly, even this should not be a problem.
In my experience, even in an overly saturated niche, there will always be enough long tail keywords available to build a site around. Every day new products and technologies are being introduced to the market, which in turn produce new search terms that people are typing into Google, which in turn makes Google starving for that newer content. It’s never ending.
For our particular strategy in trying to build an Amazon affiliate niche website, the ultimate goal of our research is to find ultra-low competition keywords. We’re going after the long tail phrases that no one else wants and because we are targeting starved content, Google should rank us in the top 10, even with no back links. Wer’e going after the stuff that the other entrepreneur’s in the industry deem too low to worry about. That’s our gold, our piece of the pie.
Keyword Research Process
I am currently using four different tools for my keyword research process. Each have benefits that others do not and are important to process our initial keyword raw material into the fine gold well be building our site around.
- Google Adwords Keyword Research Tool – Initial Grunt Research Work. IGRW.
- ILovePageOne – Used to get “AllInTitle” data, using proxies.
- LongTailPro – Used for the Competitive Analysis (only available in the pro version)
- Google Excel Spreadsheet – My own creation and used for final filtering and management. I will be sharing this in the next niche keyword research tutorial.
There are currently five steps in the keyword filtering process we will use, each step will future refine our list. Unfortunately we will not know the validity of the niche we have chosen until the very end. This process usually takes me about 2 days if done right. Longer if these tools did not exist.
Step 1. Keyword & Phrase Seed Words
We need to first generate lists of raw keywords to work from. We will be using the Google Keyword Research Tool for this stage.
The first thing you need to do is think of a list of keyword & phrase seed words. You will end up putting these into the Google keyword tool to start the process. To get ideas I will usually visit the Amazon Best Sellers list again for the product theme I am targeting to get some ideas. Try to think of words that have intents to buy.
Here are some example seed keywords for our niche example -“Water Filter”:
- …. And So on
Once you have brainstormed some seed keywords, it’s time to start generating the raw keyword lists.
Step 2. Generating The Initial Keywords Lists (IGRW)
Open up Google Keywords Tools:
Next, add in our main niche focused keyword “water filter” and one of the brainstormed keywords. Then search.
On the results page you will get ALOT of results, some specific and some not. We need to narrow down this list and make it more specific for “water filter brita”. To do that you will need to include the brain stormed keyword in the “Keywords to include” section, as seen in the image. This will ensure that your result will be specific for this keyword phrase.
You could just use the first results for “water filter”, but in my experience you miss out on a TON of keywords that this tool will not always show in a generic search.
After refining our search we are left with about 164 focused keyword phrases (out of about 800 initial results)! Now it’s a matter of downloading these results as an Excel CSV to your computer for the next step.
Repeat this step using your main keyword along with a seed word. For example the next list I would generate would be water filer best.
Step 3. Get the “AllInTitle” Data
By this stage I will have about 20-40 different focused CSV keyword result lists from using the Google Keywords Tool. The next step is to filter and find the AllInTitle results for each keyword.
This step is important for the type of strategy that we are employing. Other strategies tend to branch from this point. Basing a niche site around the AllInTitle data is just one strategy to follow, there are many others.
For this step we are going to use a keyword research tool called iLovePage1, which allows us to add in pre-purchased proxies for our AllInTitle checks.
LongTailPro is another fantastic research tool that can do most of these steps all in the same program, the problem comes when trying to find the AllInTitle metric. At the time of this writing LongTailPro does not have the ability to include searching by proxies.
Before we start, a quick refresher on the AllInTitle metric. AllInTitle is a Google search technique that targets keywords in the page titles. This can be accomplished by searching in Google with query “allintitle: brita water filter review“. The number of results are the number of pages that have this unique search phrase within their title, which does not have to be in order. The theory is that Google still ranks this as a high SEO factor, especially if there is low competition and high search volume for that keyword phrase. This is one of the reasons why its centerfold for our case study.
The problem with searching for AllInTitle metric data is that Google is starting to hammer down on this kind of search. To the point where they have purposely increased the search time AND may even ban your IP address for a few hours if you do too many. You will also get a lot of captchas.
SO, this is where the keyword research tool iLovePage1 comes in. This program has the ability to use proxies which will not only provide numerous running threads checking AllInTitle results, but will be from locations other than where you are working. This will increase the speed of this metric search and ensure that the IP address at your location will not be banned. There are problems even with this, but this post is already too long to go totally in depth ; ).
I wont go into depth with how to use iLovePage1, there are fantastic tutorials on their site that cover that. I will just be going over the important steps for our purposes.
Once you have set up the proxies in iLovePage1, add your first Google CSV.
Once your CSV is loaded, set the keyword filter phrase length to three words, along with max monthly searches at about 1000. For our example of using “brita water filter”, this reduced our list from 167 to 134 results. Once you have done this, you’re ready to start analyzing.
Depending on how many proxies you have set up this can take a LONG time. For me, with about 20-40 Google Keyword Tool results lists (each list having between 100-1000 keywords to check), this will take me about a day to analyze. The more proxies you have the less time. I usually use about 20 dedicated proxies.
Once the AllInTitle results are in, its time to filter again. We want to remove all keywords with an AllInTitle result of 1000 or more. By the end we have whittled down our initial list from 167 to 64 results. These 64 results will need to be furthered filtered in the next couple steps. For now save this list in a new CSV file called “processed_brita”. .
Continue this process until you have processed all the Google keyword raw csv lists. In the next tutorial well cover additional filter and keyword management.
This ends part 1 of our niche keyword research tutorial. We still need to cover the last couple parts of this process which involves getting the overall keyword/phrase competitiveness and then final filtering in a very cool spreadsheet to see exactly what we can use and what we cannot.
What do you think? Anything you would do differently?