Keyword Research for Adsense Sites
Wednesday, May 19, 2010 9:46It seems that everyone has their own system for doing keyword research, and I have learned a lot from reading the strategies of others. In all honesty, I can’t even remember how many different people have influenced my current system – but reading threads from people like Steve Crooks and the Xfactor on the Warrior Forum have definitely influenced this strategy (although I maintain all my mistakes are my own!). However, when I recently evaluated the performance of my past AdSense sites, it became clear that this is something I need to do better if I want to reach my goal of $100/days.
Here is my current system and strategy for doing keyword research. Comments and suggestions for improving it are welcome!:
- Look for keyword terms that have 3,000 exact searches a month. Currently I use the MicroNicheFinder to do these searches, but if you don’t have it, you can use Google Adwords Keyword Tool (it’s free). I will, however, make an exception and go for terms that have a high advertising payout. For terms with over $10 CPC, I’ll go down to as little as 1,000 exact searches a month.
- Choose terms with at least $1 average Cost Per Click (CPC), but preferably over $5 CPC. (With Google’s keyword tool, you can choose “Show Estimated Avg. CPC” to see these numbers). You should know, however, that the average is not what you will be paid – in general you’ll get about a quarter of that amount per click, and often less.
- Choose keyword phrases that have less than 40,000 exact search results, and ideally less than 10,000 exact search results. To find what the exact search results are, you simply type in the phrase into Google using quotes. There is some debate about whether this step is necessary or not, since your main competition is going to be those sites on page one that show when you do a search without quotes. But in my experience, it is easier to rank for terms that have few sites targeting that exact keyword phrase so I include this step in my keyword research.
- Look at the SOC (strength of competition) in the MicroNicheFinder and see if it is green and (ideally) under thirty (under ten is best!). This is another questionable step, as many think that the SOC is unreliable. I won’t necessarily discount a keyword term if it yellow or red, but I’ll be much more skeptical about it when doing the other steps.
- Head over to Spyfu.com and check and see if there is a good advertising base for that keyword phrase. Generally I want at least seven or eight advertisers bidding on that keyword term, and the more the better.
- Check the top three results for that keyword phrase without quotes. My goal for most keyword phrases is to get into the top three, as I have noticed there is a huge difference in revenue once I hit one of those top slots. I look at both PR and number of back links a website has, plus the on page SEO. If there is a slot that has PR2 or less, has under 500 backlinks, and/or does not directly target the keyword phrase I’m trying to rank for, I will consider it.
- Check to see if there is a .com, .net or .org site available for that keyword term. I prefer to get an exact match domain (for example, if I am targeting “automatic litter boxes” I would want automaticlitterboxes.com), although if I think the term is really good I might consider either adding an extra word (automaticlitterboxes4u.com) or dashes (automatic-litter-boxes.com). However, I really try avoiding doing this, as in my experiences domains with dashes and extra words take more work to rank high.
- If a term looks good, then do a search for secondary keywords as well. In my experience, I will often start ranking for long-tail secondary keywords first if the main keyword is pretty competitive, and sometimes you can find real gems in this process.
Right now I’m creating a spreadsheet with possibilities in it, and in June I’ll go through it and pick the best ones to build. I might add, that this is primarily for AdSense. For ClickBank and Amazon affiliate websites, I often start with a product first that I am interested in promoting, and then try to find keywords that work with that product page.
Related posts:







Tinaman says:
May 19th, 2010 at 8:38 pm
My steps are almost the same as you. But there is another additional step I took. I check against Google search based keywords tools http://www.google.com/sktool/#. You will notice the results are vastly different form MNF and adwords tools. If the keywords I choose meet all category you mentioned, I will do a check at sktool. I will forgo any search result that is lesser then 800. But saying that, not all KW are born equal. Sometimes you will have a killer KW that looks like a flop that can score home run, vice versa.
[Reply]
Michelle Reply:
May 19th, 2010 at 8:42 pm
Wow, I just checked out the tool and can’t believe I’ve never used it. That’s a great tip – Thanks!
[Reply]
Mike Roosa says:
May 20th, 2010 at 9:41 am
Most importantly though is to test, test, test. If < 10,000 exact search isn't working for you then drop it to 5,000. You feeling me?
[Reply]
Michelle says:
May 20th, 2010 at 2:36 pm
Sure, Mike, it’s always good to keep in mind that your need to adjust your strategy if it isn’t working!
[Reply]
Ruth - Web Career Girl says:
May 21st, 2010 at 11:36 am
Thanks for sharing your strategy. I must admit I am one of the ones who pays less attention to SOC and search competition numbers, and more to the PR/ backlinks of all the sites on page 1. But I still don’t feel that experienced with keyword research, so I really appreciate all these tips. I’d never seen that tool Tinman suggested – not entirely sure how to use it though, so if you’re reading Tinman, could you explain?
Anyway, really appreciate posts like this, they are very helpful!
[Reply]
Yolanda @ Flippa Chick says:
May 22nd, 2010 at 2:02 pm
I also use the http://www.google.com/sktool/# keyword tool! Actually, I found an awesome niche that I plan to turn into an authority review website based on this tool because it gives you a more realistic outlook on what’s really being searched.
Also, do forget to match up whatever data you find with Google Trends!
[Reply]
Michelle Reply:
May 22nd, 2010 at 3:21 pm
Ah, yes, that’s a good idea!
[Reply]