I’ve published my thoughts on keyword research for new sites numerous times; but this is the first time that I’ve decided to share my way to do keyword analysis for a well established resource.
Why is it difficult?
It is hard to say whether it is more (or less) difficult: to analysis keywords for an established site or for a new one. It is just different. I am not talking of the case when you are developing the site from the very start in which case you are likely to be doing keyword research weekly when checking your analytics for new search referrals.
I am talking about the case when you are asked to optimize an established site and this is the first time you’ve seen it. In this case you don’t know how it’s been evolving and you can hardly tell what words have a good (missed) opportunity to drive highly targeted traffic.
Where to start?
Unlike with a new site where you usually start with the general niche research and then broaden it with keyword suggestion tools, keyword analysis for an established site should be started with its internal analytics, of course.
The first problem you are likely to face analyzing search referrals for a well-established, high-traffic site is the overwhelming volume of long tail which is almost impossible to figure. So the steps you might want to consider are (I am describing how to do that with Google Analytics but any advanced analytics tool can do that too):
- Look at the top referrals first. Don’t forget to set a longer period of time not to miss some good words the site used to receive traffic in the past.
- Try to sort top referrals into patterns. I mentioned how to do that in the last week’s post on discovering your most frequent modifiers. What you need is to determine your core terms and most frequent modifiers for each one. The best tool to do it is Excel as, among other options it provides us with, it also enables to visualize those patterns with different colors.
- Organize your keyword phrases and patterns by topics, levels and modifiers. Again, Excel is the best tool for that. The more time you spend on sorting your keyword list into tables, the better idea you will get on how to better manage your future keyword optimization strategy.
- Organize your words by topic: general versus specific; informative, navigational, brand-specific – the topics depend on the site and your goals;
- Organize your words by levels where basic levels are determined by your core term and frequent modifiers;
- Organize your words by modifiers: this is basically meant to make sense of your patterns you determined in step 2.
- Research your current rankings: comparing your current keyword position to traffic numbers will give you an idea how promising each keyword is. For example, if your site gets 500+ uniques weekly for a word that ranks on page 2, you can safely assume it has a huge potential to send a lot of traffic when pushed to page 1. Rank checking in this case aims to answer two questions:
- How much traffic a page is likely send (with some deeper research, by looking at bounce rate and visitor’s paths, you can also predict how much conversion potential the phrase has);
- How easily it will be to achieve page 1 rankings for the given phrase (if your page has been on page 2 or 3 for the phrase for some period of time longer than one or two weeks, you stand very good chances to achieve page 1 rankings if you focus some effort on it).
- The above steps will make you remember the site keywords and patterns by heart. Now it’s time to do some third-party analysis (keyword suggestion tools and/or competitors research).
Note, that any of the above steps can be enhanced by testing keyword behavior with PPC.
SEO tools and articles to help you organize and manage your keyword lists:
- Organize your keywords with WikiPad (free);
- 2 desktop tools to help you manage your keyword lists: Traffic Travis ($100 with free basic edition) and Web CEO ($199 with free (limited) trial version);
- Organization is the #1 tip for PPC (by smart Kate Morris);
- Concentrateme: a new way to track your long tail ($35 per month).