3 Steps for Optimizing Content for Long Tail Keywords

The following is a guest post from Tom Demers.

One of the most pivotal aspects of driving large volumes of search traffic in most verticals is effectively targeting long tail keywords. While ranking for competitive phrases and developing link authority are certainly crucial aspects of SEO, much of ranking on long tail keywords is properly targeting and optimizing for them. A while ago Aaron made the following image as a conceptual example of how the relevancy algorithms may differ for different types of keywords:
Long tail keyword ranking factors

This article will outline a three step process for targeting long tail keywords.

Step 1: Build a Basket

The first (and possibly most important) consideration is determining which keywords to target. For this I think a three-step process is best:

Traditional Keyword Research

It’s always a good idea to do some idea generation and to get a feel for the possible variations of your specific targeted keyword by utilizing a keyword research tool. For the sake of the article, we’ll assume that we’ve selected our “head” or core keyword target, and that we’re attempting to rank an article for the key phrase and related key phrases. Three tools that I find particularly useful for this purpose are Google’s Search-Based Keyword Tool, the SEO Book Keyword Tool, and my company’s Free Keyword Tool.

Using Your Own Analytics

Really the best source of keyword data for determining the long tail keywords you can target is your own data. This is powerful because it shows you a variety of keyword combinations, the data is proprietary (your competitors didn’t pull the list from the same keyword tool you used, so they won’t be targeting the same keywords), and you have actual data both that you can rank for a given keyword, and you have an indication of how that keyword performs on your site. In Google Analytics, there a couple of reports you can pull to get this information (most analytics packages will provide you with similar capabilities). Drill down to traffic sources > keywords > non-paid:
Long tail keyword content stratgies
Then you can create a filter for the head term. For the sake of this example we’ll say we’re targeting the phrase “long tail” and variations:
Long tail keyword filter in Google Analytics.
By creating the filter, we can see a variety of modifiers that the page and/or other content on our site are already driving. And, if we are in fact attempting to optimize an existing page for multiple keywords, we can utilize a content report to see what that page is already driving traffic for:

View Entrance Keywords for a page in Google Analytics..

You can then see all of the queries driving traffic to that page. By analyzing the traffic and conversion statistics for that page, you can then start to feature more effective variations more prominently. The beauty of analyzing your own data lies in the fact that you can de-emphasize variations that don’t convert for your site.

Continually Iterate on Both Keyword Research and Keyword Analysis

Periodically, it’s a good idea to return to traditional keyword research, and to dig back into your analytics. This is particularly true if a concept or product is seasonal, but regardless the queries driving traffic to your site are bound to shift, and analyzing both the segment of keywords you’re targeting and the actual traffic to a given page can help to drive a tremendous amount of additional traffic to an individual page.

Step Two: Put It On The Page

Unless you coordinate an army of writers or build a venture-backed model around creating a piece of content for every phrase imaginable, you can’t create a piece of content for every phrase you want to rank for. As such you’ll have to effectively target long tail keywords by including the multiple phrases in your keyword bucket throughout the page:

  • Varying the Title Tag and Header - In varying title tags and headers for SEO you are ensuring that your pages aren’t over-optimized and they include relevant long tail keywords you’ll want to target (rather than redundantly featuring the same keyword twice).
  • Place Variations and Modifiers in Your Content - By researching the variations of a keyword you might want to include in your content, you can be aware of them as you craft content, and you can strategically place modifiers throughout your page’s content. For instance, it might not be natural for you write out the phrase “affiliate long tail keywords for promoting products” but if you know this is a phrase that drives some traffic, you can be sure to include phrases like “whether you are a retailer or an affiliate promoting products”. You’ll be using phrases like long tail keywords frequently enough that if the longer phrase is lower competition, you might not even need to include the exact phrase to rank for it. Note below that none of the ranking pages use the exact phrase “affiliate long tail keywords for promoting products”:
  • This is the SERP for affiliate long tail keywords for promoting products.

  • Pay Attention to All of Your On-Page Elements - Be sure to work into your page’s headlines, bolded copy, alt attributes, title attributes, etc. the variations you’re targeting. By mixing up the words and phrases you use in these elements, you’re also ensuring your page isn’t over-optimized

Step Three: Building Links For Your Keyword Basket

Finally, even though many of your long tail keyword variations will rank on their own, you’ll want to develop some links with specific anchor text to these pages. You can do this in a few different ways:

  • Vary Your Internal Links to a Page– Again, this allows you to avoid being “over-optimized,” and if you stick primarily to variations that contain the head keyword within the variation and append modifiers, rather than synonyms, you’re consistently transferring relevance for your core term.
  • Use an Important Modifier in Your Headline – While your title tag is what’s seen by searchers, many people linking to your article will use your headline as anchor text. Using a variation here helps attract links for important modifiers
  • External Links You Control- Things like company listings, directory listings, and nepotistic links often offer you the opportunity to control your own anchor text: while many times just leveraging internal links on an authoritative site is enough to rank, sometimes utilizing article submission Websites or other low-quality external linking sources with keyword-rich anchor text can help you to rank for mid to low-competition keywords.

Ultimately the best way to rank for long tail keywords is to build an authoritative Website and seed it with a lot of content, but on a page-by-page basis you can often leverage strategic keyword targeting and your own analytic data to help drive exponentially more traffic than you would focusing solely on the “head” keyword.

Tom Demers is the Director of Marketing with WordStream, a software company specializing in pay-per click software and keyword research and organization solutions for SEO. Tom is a frequent contributor at the WordStream Internet Marketing Blog.

Published: February 22, 2010 by Aaron Wall in


February 23, 2010 - 10:36pm

It seems that the focus on analytics data for long-tail research is somewhat flawed. To me, the scalability of the long-tail is what makes it valuable. A keyphrase/landing-page level analysis is not scalable at all. Long-tail targeting is really not a complicated venture.

1) Identify core keyword(s)
2) Build deep, relevant, unique, and useful content surrounding keyword(s)

The example of “affiliate long tail keywords for promoting products” demonstrates perfectly the flaw in the logic used. As explicitly stated, the top ranking sites are not specifically targeting this keyword and yet are ranking well. If the site owners were to look into their analytics data and identified “affiliate long tail keywords for promoting products” as a high-converting, long-tail keyword for a specific landing page, your recommendation would be modifying that page in some way to better optimize for that keyword ("By analyzing the traffic and conversion statistics for that page, you can then start to feature more effective variations more prominently.").

These sites already rank well for this term, despite no overt attempt to rank for the specific keyword. The effort required to further optimize for this keyword is likely not worth it due to the very fact that it is a long-tail (low-volume) keyword. And more importantly, is not scalable at all. The site-owners in this example would be much better off continuing to build unique, deep content relevant to the head keywords. The long-tail is not about splitting hairs for specific keywords, it's about most effectively capturing a broader range of keywords (and thereby a greater volume of overall traffic).

Additionally, by focusing so much on keywords that are already driving traffic to your site, you could very well be missing out on bigger opportunities for new markets that are not being captured. While your own site's data is "proprietary", it is also a very siloed way of looking for keyword expansion. This becomes all the more dramatic when considering long-tail opportunities. The long-tail is your opportunity to perform extensive competitive/market research and identify head keywords that you may not be able to rank for, but may be able to build a content strategy in order to capture related long-tail keywords.

February 24, 2010 - 12:42am

At first, I was opposed to your point, because I agreed with the ideas in the post - but your argument was very well stated.
"The long-tail is not about splitting hairs for specific keywords, it's about most effectively capturing a broader range of keywords (and thereby a greater volume of overall traffic)."
I don't think these ideas are separate, or mutually exclusive. The long-tail is about splitting hairs on specific keywords, so you can more effectively capture a broader range of keywords. It is about analyzing the traffic, and optimizing the efforts to convert it. This means seeing oppportunities in the analysis, and building for it on future pages.
But you make some good points about the scalability - it can easily spin off into busywork if a keen eye is not kept locked on the ROI. I would not try scaling it, as much as focusing it.
I don't agree that most small business are best suited by targeting harder keywords, and hoping for the associated long-tail trickle-down theory to happen naturally. Instead, I would think they use their analytics to see where consistent conversion patterns exist, and then weave in longer tail strategies to strengthen this association for the domain.
A long tail, specifically tailored answer to an equally specific query is a page that is not usually designed for receiving boat loads of traffic. It is designed to have a significantly higher conversion rate.
There is a big difference in the approach there, and I think Tom's approach would lean more toward mine, where you would lean more toward increased traffic. The balance between more traffic and conversion rate is the eternal balance - I guess we are all looking for increases in "refined" traffic.
But thanks for the great comment LRath, and making me think about it harder.
And thanks to Tom and Aaron for the interview too.

February 24, 2010 - 2:42am

@lrath and @martypants: outstanding comments! Thanks for weighing in.

Lrath you make some excellent points, let me clarify a few things:

  • I did make leveraging your own analytic data more of a focus, and while I gave it a quick mention I probably should have focused more on identifying secondary/long tail keywords outside your own data (traditional keyword research).
  • I see your point about the authoritative sites ranking "by accident" for that term, but if it is a term that converts isn't it better to "steal" that ranking from them by being aware of the term? Additionally my overarching point with that example (which admittedly may not have been the best one, in hindsight) is that you should be aware of the terms that drive traffic and convert as you create your content. Just as you would look at your internal and external data to determine which "head" keywords to target, you should also know the granular variations and modifiers a page should be targeting as you create longer content or optimize existing content - often the aggregate of the long tail is actually greater than the head terms on a page, so significantly increasing traffic on your entire portfolio of long tail terms by being cognizant of variations and modifiers can actually have a huge impact. Why would you just target the head terms, write a lot of long-form copy and leave the rest to chance when it's relatively simple to have a good understanding of the variations and their relative value?
  • I'd also echo Aaron's answer here and say that often times the best way to see what you can actually rank and drive traffic for is to publish a page and monitor the traffic: as MartyPants said not everyone can/will rank for the terms the Google says they should, and often times a business might spend hours and hours and large sums of money optimizing for a term that they may find out in the end won't actually convert for their business (this is a benefit of the "run PPC first for keyword research approach").

All that said I'd agree with your overarching points that building an authoritative site with a lot of keyword-focused content is paramount, and I'd also agree that a major part of keyword targeting/expansion is traditional keyword research, which I should have better emphasized.

Thanks a ton for the great comments, they both really added to the post; hopefully I clarified a bit here in the comments.



February 24, 2010 - 4:01am

Completely unrelated comment... sorry, I'm an idiot and don't know many influential people.

eBay does not verify new accounts adequately, which opens the doors all kinds of malicious behavior, namely false bidding.

I've had multiple cases where 2 or more, zero feedback, bogus accounts bid against each other for my items, driving the price up and legitimate bidders out. I then have to relist the item, cancel bids and block bidders.

But eBay benefits from this to an extent, so they aren't doing anything about it. If I don't report an item as unpaid within the allotted time, eBay keeps the final value and insertion fees. This also has an inflationary effect on pricing, which I suppose could be a good thing for the few sellers who aren't victimized. Furthermore, eBay would prefer to keep the new account registration process as simple as possible in order to increase sign-up rate.


If you blogged about it, and tied it into behavioral SEO theory... CTR and Bounce Rate authority given to iGoogle account holders who, by the fact that they have multiple email exchanges, long search histories, credit cards on file, etc, prove that they are real people and their online behavior can be trusted as reliable data to be mined... or something, eBay might do something about it.

Or not. Like I said, I'm a presumptuous idiot.

February 24, 2010 - 12:24pm

Hi Direct,

I think a more interesting parallel here is comparing the above behavior to things that happen on AdWords, and particularly the content network; not sure I'd write that (and even if I did this was just a guest post; I'm not a regular blogger here) but I'll keep it in mind, thanks for dropping by!


February 25, 2010 - 3:21pm

Thanks so much -- all great tools I use every day.

February 25, 2010 - 5:52pm

Using your unique analytics data as shown above can be good for:

a. phrases that you have no page optimised for; where
b. you rank somewhere like 10-30
c. and users found what they wanted on your non-optimised page (and they didn't on the sites on page 1) and stayed around on your content for a while.

This is a good indicator that a bit of keyword optimisation might get a new optimised page a good ranking position and you will get a fair amount of very relevant traffic from it.

PS Since Aaron did his original diagram has he had any cause to change the % estimates for each slice of the pie chart?

February 25, 2010 - 6:48pm

Those percentages were *really* rough, sorta more as a way of helping set up a mental model for keyword rankings, but they did not aim to be in any way precise.

Not only do the relevancy algorithms change (by search engine AND over time), but each keyword market is unique (with different sets of opportunities), and there is no consensus as to what "long tail" and "head" keywords mean (sorta depends on how broad and ambitious a person is).

Based on that I don't really have any precise suggestion as to the exact amount anything counts without at least seeing the market and doing a bit of keyword research. The above was just a really rough rule of thumb.

Melissa Gonzalez
February 26, 2010 - 7:55pm

Nice Visuals, thank you so much for the post and the links on the keyword tools. I had no idea SEO Book had one. I really like using analytics as a backup to my keyword research. Its always interesting to see if your content writing efforts for the long tailed keywords are working and what exactly it is that people are typing in to get to you.

February 28, 2010 - 4:24am

I agree with the notion of looking at the keywords in our analytical data because a lot of times it seems as though we discount those keywords as if we "ought to be getting traffic" from those anyway.

Let's say you're optimizing for a broader keyword, but a longer tail is driving some good (i.e. higher converting) traffic to the site, why dispute that? Your audience is trying to tell you something.

Sometimes we ignore the obvious in an effort to appear more "intellectual" or to "out-think" our visitors. Why?

March 4, 2010 - 5:57am

WOW! Valuable information. We concentrate in keywords but completely forget about long tail. I will site here for a while and read your post again. But I should also reconsider my SEO strategy. Thanks a lot for this posting!

July 8, 2010 - 4:01am

Trying to rank on page number one of Search Engines for highly competitive keywords such as ‘gym’ or even ‘Sydney gym’ can be difficult but by analysing your business, you can break down the keywords into longer tail keywords that your target market may use to search for your business.

So instead of targeting the page for a trophy keyword like ‘gym’, you may want to target the page for a niche phrase like `women’s gym eastern suburbs Sydney’ or ‘personal trainer women’s gym eastern suburbs Sydney’ and be in the running (pun intended) to obtain traffic from long tail keywords combinations.

Highly specific multi-word phrases tend to be far easier to rank well for than more generic single keyword or double keyword phrases. However, the traffic will be better targeted as people who search with long tail keywords are far more likely to become buyers. Remember, the end goal of SEO is not to increase traffic but to increase your number of conversions.

March 27, 2013 - 8:19pm

Aaron and Tom, would you say this is still accurate as far as long tail keyword optimization is concerned?

March 27, 2013 - 8:48pm

The above still generally holds true, but things have changed somewhat in the past 3+ years.

A couple things to note:

  • The above article was published a year before Panda, & Panda impacts the latent risk embedded in generating lots of content pages.
  • The above article was published before the disavow tool was created & article directories have certainly fallen out of favor with Google.
November 28, 2017 - 11:08am

getting exact keyword link from article submission sites still worthy in 2017?

December 1, 2017 - 11:42am

...have a combination of a broad topical selection, low quality article submissions done just for the link & an ad heavy layout.

That combination in turn has led to Panda-related penalties. That in turn led to declining traffic & ad revenues, which in turn led to botched "clean up" efforts where all outbound links were nofollowed, or sites were sold & sold again (like ArticlesBase).

Even About.com was spun off into a number of vertical sites.

The short answer is the sites which will be online for years to come are probably those which have some brand value, a unique editorial voice, cater to a specific niche, etc. ... mostly the exact opposite of broad article directory sites.

There are some vertical community sites which publish guest articles. Some (like Forbes) are really spammy with the ad layout and format. Others (like SeekingAlpha) may be a bit aggressive with pushing for user registration and using stuff like pagination, but they will certainly last longer than the broad topical article directory sites.

Many news sites that accept user submissions (like The Huffington Post) also slap nofollow on author bio links to ensure people writing on those sites are doing it for the exposure and not the link.

Add new comment

(If you're a human, don't change the following field)
Your first name.
(If you're a human, don't change the following field)
Your first name.
(If you're a human, don't change the following field)
Your first name.