Tweet
There are two physical sensations that are at once difficult to describe to anyone who has never experienced them, and yet, quite universal to anyone who has. One is the sense of urbane anticipation that comes from silently flicking your hole cards to request a "hit" at the blackjack table, and the other is the surprisingly solid weight of moving a quality chess piece across the board.
I first learned to play chess when I was about ten, and immediately fell in love with the game. For several years, I even thought I was pretty competent, regularly beating most of the other kids in the neighborhood. When I left the playgrounds of elementary school, however, for the halls of junior high, I quickly discovered otherwise. My first lesson in humility came at the hands of my algebra teacher, who not only beat me in about twelve moves, but did it while carrying on an animated conversation with half a dozen other students. I swear, to this day, Mr. Kirsten never more than glanced at the board before launching devastating attack after attack into my defensives. He taught me something important that day, and it wasn't just about chess.
Knowing something and mastering something are two very different things.
I don't much go in for hype, and heaven knows the title to this article smacks badly of it, but I honestly believe the title is nonetheless accurate. There really is a Secret to SEO. And I'm going to tell you what it is. I doubt it's going to set the world on fire, however, because knowing the secret probably won't give you immediate mastery of search engine optimization. Knowing and mastering are two different things.
The Definitive Secret to SEO is ... keywords.
I know. It sounds so simple, doesn't it? Well, learning how to move a knight or bishop is simple, too, at least until someone puts a can of whoop on you like my algebra teacher did. There are complexities in them thar hills, my friend. Let's start at the beginning ...
----- What Are You Selling? -----
This is probably the most important business question you will ever ask yourself. Unfortunately, it's also often one of the most complex to analyze and the most difficult to answer. Historically, business owners have been warned that there are two different levels to the question, both of pretty obvious importance. Harley Davidson sells motorcycles, and that's one level, what one might call the surface level. When you're building factories or contracting with suppliers, it's an important level to understand.
Beyond that, however, Harley doesn't really sell vehicles. It sells prestige and status.
That distinction is of paramount importance when it comes time to actually sell the product to a living, breathing human being. The customer, of course, doesn't walk into a store and ask to take some status on a test drive. The customer "thinks" he wants a motorcycle, but if that's all you try to sell him, you're badly missing the boat. Your sales pitch, the company literature, the image projected in advertising - everything you do has to sell what the customer is really there to buy. And that ain't just a bike.
This is old news, of course, available in any good business book. As a netrepreneur, you still need to understand what you are selling. You need to differentiate between what the customer thinks they are buying and what they actually want to buy. You will then make or purchase the former, while all of your sales and marketing efforts will be directed at the latter.
The Web, however, adds a third level to the question.
This third level exists where the line between the other two levels becomes blurry. It exists because, on the Internet, people have to FIND YOU before they can buy from you.
Technically, this is equally true in the brick-and-mortar world and is a problem faced by every business. People still need to find your business. One solution, often an expensive one, is to rely on branding and advertising. Another solution, one especially used by local businesses, is something called cross-advertising. The smart businessman doesn't just buy space in the Yellow Pages, but rather buys space in multiple categories in the Yellow Pages. Different people will look for a business in different ways.
The problem of people finding your site becomes a third level, of equal importance to the other two levels, because of the unique nature of the Internet. Unlike the Yellow Pages of the brick-and-mortar world, alphabetical categories for a business just don't work very well. People can still find your site through branding, through frequent advertising, even occasionally through browsing. But most of the people have become much more accustomed to SEARCHING for what they want on the Web.
Let's say you wanted to buy a Harley Davidson on the Internet. How would you find one?
Chances are good you would go to your favorite search engine. If you enter "motorcycles" as your search term, however, you're going to be quickly inundated with several million irrelevant results. Even though that is the first level answer to our "What are you selling?" question, it's far too broad a keyword to get the buyer to where they want to go.
The second level answer is even more useless. If they enter "status" or "prestige" (which we know is what the customer really wants to buy) it's highly unlikely they'll ever find a site to buy their Harley.
The third level answer to our question is usually where the other two levels meet. They might enter "expensive motorcycle," or maybe "motorcycle sportster," or perhaps even "American motorcycle chopper hog."
NOW, we're talking keywords -- the words or phrases people actually use to search for something.
(Yea, I know the first search term they use will probably be " Harley." That is not where our first and second level answers merge, but rather is the result of very effective branding. It's important, but doesn't help you much unless the name of your business has been just as heavily branded. Someday soon, we'll talk about the importance of branding, but right now we'll continue to concentrate on the more short-term importance of keywords.)
----- Keywords are the Key -----
What are you selling? When you find the answers for all three levels to this question, you're well on your way to developing a keyword list. Just like goodwill, this is an intangible business asset.
Most people think of a keyword list in the same way a fourteen-year-old thinks of a chess game. Move the knight here, the bishop there, and it's a wham bam wanna-play-again game. You can play chess like that, and yea, you can put together a keyword list in maybe thirty minutes, but neither approach will make you a master.
Developing a good keyword list is a complicated, very iterative process that requires time and effort. But that's not a bad thing. If it was too easy, after all, your keyword list wouldn't be worth much as a business asset. The time and effort are investments that can eventually pay big dividends.
Unfortunately, to some extent, keywords fall into the chicken and egg paradox. Which comes first? You can't really fine-tune your keywords until you have an active, running business. But you shouldn't even think about starting a business until you have chosen your primary keywords. Conundrum city!
However, while you can't fine-tune or finalize your keywords you start using them, what you can do is understand some basic principles and concepts. With those as your foundation, there is hope you can select at least the most important keywords for your site.
----- Building a Foundation (or, This is IMPORTANT Stuff!) -----
If we were to graph the relative importance of keywords across the Web, they would be best described by a standard bell curve.
At one end of the spectrum are the keywords that are so highly competitive as to often be useless to you. Go to Google and search for "travel." Notice all the sponsored links? There might be a few government or educational sites (.gov or .edu), but you'll probably notice that most of the results are extremely well known, very long established web sites. Unless you have a few million dollars in the bank, I wouldn't recommended trying to compete with Microsoft or Travelocity for this particular keyword.
This side of our bell curve is filled by the words and phrases that are the most searched on the Internet. They include travel, mp3, jobs, sex (of course), music, food and many more. Millions of people enter these words into a search engine every single day. And thousands of web sites are trying to capture that very big audience. It's a bit like the lottery. The rewards for winning seem gargantuan, but your chances are abysmally slim. You compete for these keywords at your own risk!
Later, I'll offer some tools and references to help you recognize the highly competitive keywords. Much later, on another day, we'll also talk about ways you CAN learn to compete with the big boys with a margin of success. After all, you can't win the lottery unless you at least buy a ticket.
(Did anyone notice I switched from selling motorcycles to selling something in the travel industry? The best keyword for our hypothetical timepiece site would still include "harley," and I wanted to avoid complicating this discussion with trademarks.)
At the other end of our bell curve are keywords that are so highly targeted that John Q. Public will never think to use them when searching. The keyword "peregrination" may be a delightful synonym for travel, but it won't bring you a lot of visitors.
In the middle of our bell curve lies the bull's eye. These are the words and phrases people actually will USE when searching for products or services. And that brings us to another, albeit slightly different, way of looking at the same thing.
----- Painting a Target on Your Customer -----
Contrary to popular opinion, you don't really want visitors. You want customers.
If you could build a web site that came up on the first page when someone searched for "britney spears," you might get a ton of visitors. But you'd be a bit daft to think you could sell many Harley Davidson bikes. The traffic you received would be very untargeted and your bandwidth costs would probably far exceed the occasional and random sale.
Selecting the keywords for which you will optimize your web site is how you prequalify your traffic. Say that three times, real slow. Let it sink in. It's important.
If you select keywords that target your traffic, your conversion rate (how many people visit versus how many people buy) should be fairly high. But, uh, not too high. If your keywords are TOO targeted, you're almost certainly missing a large chunk of your potential market. Gee, that sounds a bit like a curve, too, doesn't it?
It's not really a different curve, though - it's a subset of the same one we talked about previously. The subset includes only the keywords specific to your product or service. At one end are the keywords with the potential to bring hoards of largely untargeted visitors to your web site, at the other end are the words that will bring a very few highly targeted people, and in the middle are the ones that represent an equitable ROI (Return on Investment is something we'll discuss in just a minute).
Unless you're selling U.S. dollar bills for eighty cents each, you will never want to promote keywords too far to the left on the curve. This is a critical point to understand, so let's talk about it for a minute.
----- Using ROI to Select Your Keywords -----
Simplified, we can say Return On Investment = Profit / Cost * 100.
Let's say you get 1,000 visitors per day to your site from the keyword "e-commerce," and it costs you one U.S. penny for each one. That cost comes from several different places, but we don't need to worry about those yet. You just invested $10 USD. Unfortunately, these are very untargeted visitors and you manage to sell only one e-commerce widget, for $5 USD. Your Return on Investment (ROI) is -50%, resulting in a net loss of five bucks. Oops.
Generic keywords can bring lots of traffic, but the traffic is very poorly qualified. They don't stick around your site, and they rarely buy anything. They are window shopping, plain and simple. That doesn't mean you don't want them, of course, because they do have potential value down the road. It just means you don't want to spend all your resources attracting them and, because these terms are often highly competitive ones, they DO require a lot of resources.
Increasing your search engine rank for the keyword "e-commerce" won't help!
Getting 10,000 equally untargeted visitors every day only means you lose $50 instead of $5. When you're losing money, you can't make it up on volume. A negative ROI is always bad. You are too far to the left on the Keyword ROI curve and the only solution is to move to the right, to a more highly targeted keyword.
Generic keywords don't always go negative, of course. But they do invariably result in a lower ROI, sometimes too low for a business to survive.
Okay, so you make some changes to your web site copy, targeting the keyword "e-commerce widgets in Southern Michigan." You now get 10 visitors a day to your site, and it takes a week to sell one of those visitors an e-commerce widget. Those seventy visitors still cost a penny a piece, seventy cents, and you still made $5 on your single sale. Your net profit is $4.30, with a very healthy ROI of 614%.
If you can increase your SE rank for "e-commerce widgets in Southern Michigan," you should get more visitors and your ROI will likely remain the same. Ergo, more money in the cash register. However, if you already have the number one rank for that keyword (and chances are VERY good you would), you'll have to learn to live on $4.30 a week profit. You have a great ROI, but miserable cash flow. Oops.
Of course, what's happened here is that you optimized your site for a keyword too far to the right on our curve. It is highly targeted, with good converstion and a really impressive ROI, but there just isn't enough volume because very few people search that way.
So, you go back and optimize for the keyword "e-commerce widget." You get 500 visitors a day and sell two widgets. That's a net profit of $5 a day, or a ROI of 100%. Note that your ROI has dropped from a very high 614%, but you're making substantially more money. (We are grossly over-simplifying ROI by including only variable costs, but it should still serve to make the point.)
Selecting your keywords is the foundation upon which everything else in SEO must be built.
----- Search Engine Keywords -----
The secret to SEO is to not select a single keyword, or even a handful of keywords. You wants scores, hundreds, possibly even thousands. Instead of a few highly visible keywords that bring in a ton of untargeted traffic, you want a hundred less competitive keywords that TOGETHER can bring in almost as much very targeted traffic.
So, how do you choose the right keywords. The first step in selecting the keywords for your site is to make a whole lot of educated guesses.
Try to find your competitors using just a search engine. Keep good records of what keywords you try and what works best. Note the rank for each keyword for each competitor. Did they come up number one on the search? Or were they buried on page nine? A spreadsheet makes an excellent tool for this early analysis.
The only keywords that really matter are the ones that work, but you can also often get an idea of what your competitors "think" are good keywords. Most browsers will let you right-click on your competitor's page and do a "View Source." Look near the top of their source code for something like META NAME="keywords" and then examine the Contents field that immediately follows it. These are the keywords your competitor has chosen to target, and they might give you a few fresh ideas. If nothing else, you can use these Meta Keywords to fuel additional searches.
Once you've built a list of keywords, take it and set it aside.
Depending on your industry and your own history within that industry, what you used to find your competitors might be of dubious quality. There's a real good chance you know too much to be a "typical" customer.
It's equally true that your competitors may be similarly blinded by their own industry experience. You may be finding their sites only because they targeted the wrong keywords, the keywords on the far end of the curve that only insiders would typically use. If you make the same mistake, the only people who will ever find your site will be your competition!
Ask your friends, someone outside your industry, to do the same thing you already did. Have them find your competitors using a search engine, while maintaining good records of what works and what doesn't work. They should record not only what keywords were used, but the position and page of the relevant sites found. The more people you enlist, the better will be your initial keywords list.
When you compare your multiple lists (yours and those of your friends), and combine them where applicable, you'll have the nexus of a keyword list. This is NOT your keyword list, though. It belongs to your competitors. It's how your competitors can be found. If you use this list, you'll be on a par with your competition (assuming you use it well). That's cool, but we want a birdie.
----- Extending your Keyword List -----
Now that you have a starting point using people (yourself and friends), we're going to see if we can use some software tools to extend the list.
To see this software in action, so we can better understand what to expect, we'll first look at it using example keywords. Go to AltaVista ( http://www.altavista.com/ ) and do a keyword search for "poetry." When the first page of results are returned, look for the RELATED SEARCHES section to the right. This is an often unrealized and untapped gold mine!
When people go to AltaVista and search for a generic keyword like "poetry," they very often don't like the results. Too much forest and not enough trees. So they try another search, a refinement of their previous, in hopes of getting better results. Indeed, studies have shown that most searches refine their attempts two or three times before getting good results. AltaVista tracks those refinements and when it finds a refined search used by a lot of people, it offers that as a suggestion to others. So, we know that people who start out searching for "poetry" will also very often search for "love poetry" and "poetry contests."
Can you spell i-n-v-a-l-u-a-b-l-e?
We did this first with a simple example, because AltaVista won't offer suggestions unless your keyword is fairly generic. You should try it with each item in your keyword list, but don't be surprised (or discouraged) if many or even most don't return a suggestion list. That just means you already have nicely targeted keywords. Where you do get a suggestion list from AV, however, you should use those to extend your list.
If you think this kind of research is time-consuming, entering each of your keywords into AltaVista, I almost hate to tell you that you're just getting started.
AltaVista isn't the only search engine that includes a suggestion list, and you need to duplicate your efforts on a number of others. Here's a list of some that will help you extend your keyword list - but you should also realize that search engines change almost daily and this list will never be all-inclusive. Every time you use a search engine, check to see if it has been updated to include the Suggestion feature. Then use it!
Excite http://excite.com/
Teoma http://teoma.com/index.asp
Lycos http://www.lycos.com/
WebCrawler http://www.webcrawler.com/info.wbcrwl/
AllTheWeb http://www.bos2.alltheweb.com/
Getting' tired, yet? I hope not, because we still have a few more stops to make.
Some day soon in the Gazette, we're going to spend time talking about what many call the PPC (Pay Per Click) Search Engines, but we need to understand a little about them right now. There are a lot of PPC Engines, but the two largest are O*****re and Google. PPC simply means that when a visitor clicks on the link returned in a search, the web site (who has now become an advertiser) will pay the search engine a few pennies (or dollars) for the visit. How much they pay can be determined in several different ways, none of which matter to us right now.
What is important to us is that the PPC Engines want people to advertise their keywords.
To entice web sites to advertise, some are willing to tell us how many times a specific keyword has been used in their search engine. Virtually all of the PPC Engines will at least give us some good ideas on selecting keywords that are both popular and relevant. Additionally, there are a few tools available designed specifically for PPC bidding.
WordSpot, for example, offers a paid service where they will create a fairly comprehensive report for your specific keywords. A similar paid service can be found at WordTracker, along with a free trial. Google AdWords offers a tool called Keyword Suggestions that gives you a rough idea of popularity (and does offer excellent variations for your keywords), though it doesn't give a true numerical index of the keyword popularity. Like Google's tool, O*****re's will offer variations of your keywords, but it will also give a true numerical count of the number of times each keyword was entered by a human searcher. A relatively new tool, at Digital Point Solutions, brings the results of WordTracker and O*****re together for a convenient comparison.
Here are the links:
* http://www.KeywordDigger.com
* http://www.wordspot.com/
* http://www.wordtracker.com/
* https://adwords.google.com/select/main?cmd=KeywordSandbox
* http://inventory.o*****re.com/d/searchinventory/suggestion/
* http://www.digitalpoint.com/tools/suggestion/
----- Putting It All Together -----
You now have the genesis of a Keyword List. Conservatively, you should have several hundred keyword phrases, and several thousand isn't at all unusual. So far, we've been learning how to move knights and bishops around the board. Building a comprehensive list of relevant keywords is a lot of work, but it's still just work.
It's time now, to move into Mastery.
To refine your list, we should ideally look at the ROI for each Keyword you've gathered. But, uh, what if you don't HAVE any returns on your investment, yet? That's what makes this a chicken and egg paradox. Your best keywords will depend on how much profit they bring, we can't determine profit before we promote the web site, but we need keywords before we start promoting. A vicious circle, indeed.
Remember our bell curve? One side is occupied by keywords that bring a lot of traffic but not necessarily a corresponding number of sales. Most of the people on the Internet make the mistake of optimizing for one of these keywords, often making those terms very competitive. I mean, only so many sites can get a first page ranking, right? You want to avoid these keywords, at least for right now. Let someone else fight over those. Your investment will be too high, your return too low.
The other side of the bell curve is occupied by keywords that don't bring a ton of visitors, but which otherwise convert really well. These are a real goldmine, but often represent a whole lot of work (we'll talk about why in a minute).
In the middle of the curve are the keywords that bring a decent amount of traffic that results in a decent amount of sales.
Your job, now, is to go through your list of keywords and assign each of them a priority number. Make the most competitive, highest traffic phrase number one. Then find the second most competitive, highest traffic keyword. Keep going until every keyword in your list is assigned a number, then sort the list accordingly. Obviously, there's going to a whole lot of guessing. How well you guess, in large part, defines your mastery of SEO. That's the part no one can teach you.
Ready? Go to the last keyword on your list, the least competitive, lowest traffic one, and type it into Google. Visit each of the ten sites on the first page and see if you can determine why the search engine though that site was relevant to that keyword. Ask yourself, could you build a page that was more relevant? If you can honestly answer yes, if you are confident you could get a web page listed in the top ten, go to the next keyword in your list and repeat the whole process. You want to work your way up your list of keywords until you are no longer certain of success.
THAT is where you want to start optimizing your web site.
You've just found the point on your bell curve that takes into account relevance, an estimate of ROI based on competitiveness, and your own gut feelings for your current ability to get results.
Create a web page optimized for that specific keyword. Get some links to it, both internal and, if possible, external. Then, while you are waiting for it to be visited and indexed, move down your list to the next keyword you feel confident is attainable. Repeat this process until you run out of keywords.
Of course, in truth, the process is going to depend on the type of web site you're building. You can't just build pages randomly, and the structure of the site will determine the best way to proceed. However, within that structure, you need to always keep your real goal in mind.
The definitive secret to SEO is deceptively simple, just like learning to play chess. Create a bell curve, pick a point on the curve, then optimize everything to one side of that point. In all likelihood, there won't be a single keyword on that side of the curve that will make your SEO campaign a huge success. Taken collectively, however, they're going to be dynamite. Instead of getting a lot of traffic from just a few high profile keywords, you're going to be getting a trickle from bunches and bunches of keywords, each of which will bring you highly targeted visitors. Make sense?
Here's the good part.
By the time you've exhausted the keywords on one side of your originating point, you should have a really solid foundation. You've built a ton of great content, which in turn, should have brought you some good links. It's now time to return to that point on the curve and start building web pages for the keywords going in the other direction. With your new foundation, you're going to find it much easier to start attaining some of those more competitive terms.
The secret to SEO is keywords, and the secret to keywords is targeted diversity.
* Epilogue (or, Do Not Open Until Christmas) : As your SEO campaign progresses over time, you want to periodically revisit your keyword list and convert all those guessed ROI numbers into real ones. Use your server logs to track which keywords are bringing in visitors and, where possible, which visitors are converting to customers. The more you know about your keywords, the more success you can expect.
Highly Competitive Keywords
In our lead article this issue, we briefly talked about highly competitive keywords. These are the search terms that are among the most frequently used on the Internet, words like "travel" and "music." Most of the time you'll want to avoid targeting these keywords. Competition for these keywords is fierce and going head-to-head with the Big Boys is often a mistake. It will be hard to get a decent ROI because your Investment will necessarily be high to compete.
To avoid them, obviously you need to know what they are. And that's my excuse for providing this list of resources where you can find the most common search terms used across the Internet. Beyond excuses, however, it can be just plain fun to discover what interests the world this week.
Wordspot ( http://www.wordspot.com/ ) is my personal favorite. They used to provide a list of the top 200 keywords directly on their site, but have recently changed their policy and now give it only when you sign up for their newsletter. But that's okay. With the newsletter coming directly to your email box, you won't have to remember to occasionally check on their list. As with all things on the Internet, their list of words should be taken with a grain of salt. Their methodology is good, always getting better, but certainly not perfect.
CNet ( http://www.search.com/top?tag=ex.se.sr.top.top2 ) offers a list of the top 100 keywords from their own search.com utility which should again (need I say it?) be taken with a grain of salt. While this list is derived from a very substantial pool of data, it doesn't necessarily represent the Internet's "typical" searcher. CNet, after all, tends to attract the more computer-savvy user.
The Lycos 50 ( http://50.lycos.com/ ) is updated weekly to reflect the top 50 searches at Lycos. I particularly like the way they also show how many weeks a keyword has made the list, as this is obviously important information. If your product or service is seasonal, be sure to check out their archives, too.
Yahoo Buzz Index ( http://buzz.yahoo.com/ ) is very similar to the Lycos 50, but is limited to the top 20 keywords from the extensive Yahoo database.
Google Zeitgeist ( http://www.google.com/press/zeitgeist.html ) pulls from an absolutely huge database of search queries, is extremely well organized (especially if you're interested in non-English keywords), but is sadly limited to just the Top 10 keywords in multiple categories.
By Ron Carnell
I first learned to play chess when I was about ten, and immediately fell in love with the game. For several years, I even thought I was pretty competent, regularly beating most of the other kids in the neighborhood. When I left the playgrounds of elementary school, however, for the halls of junior high, I quickly discovered otherwise. My first lesson in humility came at the hands of my algebra teacher, who not only beat me in about twelve moves, but did it while carrying on an animated conversation with half a dozen other students. I swear, to this day, Mr. Kirsten never more than glanced at the board before launching devastating attack after attack into my defensives. He taught me something important that day, and it wasn't just about chess.
Knowing something and mastering something are two very different things.
I don't much go in for hype, and heaven knows the title to this article smacks badly of it, but I honestly believe the title is nonetheless accurate. There really is a Secret to SEO. And I'm going to tell you what it is. I doubt it's going to set the world on fire, however, because knowing the secret probably won't give you immediate mastery of search engine optimization. Knowing and mastering are two different things.
The Definitive Secret to SEO is ... keywords.
I know. It sounds so simple, doesn't it? Well, learning how to move a knight or bishop is simple, too, at least until someone puts a can of whoop on you like my algebra teacher did. There are complexities in them thar hills, my friend. Let's start at the beginning ...
----- What Are You Selling? -----
This is probably the most important business question you will ever ask yourself. Unfortunately, it's also often one of the most complex to analyze and the most difficult to answer. Historically, business owners have been warned that there are two different levels to the question, both of pretty obvious importance. Harley Davidson sells motorcycles, and that's one level, what one might call the surface level. When you're building factories or contracting with suppliers, it's an important level to understand.
Beyond that, however, Harley doesn't really sell vehicles. It sells prestige and status.
That distinction is of paramount importance when it comes time to actually sell the product to a living, breathing human being. The customer, of course, doesn't walk into a store and ask to take some status on a test drive. The customer "thinks" he wants a motorcycle, but if that's all you try to sell him, you're badly missing the boat. Your sales pitch, the company literature, the image projected in advertising - everything you do has to sell what the customer is really there to buy. And that ain't just a bike.
This is old news, of course, available in any good business book. As a netrepreneur, you still need to understand what you are selling. You need to differentiate between what the customer thinks they are buying and what they actually want to buy. You will then make or purchase the former, while all of your sales and marketing efforts will be directed at the latter.
The Web, however, adds a third level to the question.
This third level exists where the line between the other two levels becomes blurry. It exists because, on the Internet, people have to FIND YOU before they can buy from you.
Technically, this is equally true in the brick-and-mortar world and is a problem faced by every business. People still need to find your business. One solution, often an expensive one, is to rely on branding and advertising. Another solution, one especially used by local businesses, is something called cross-advertising. The smart businessman doesn't just buy space in the Yellow Pages, but rather buys space in multiple categories in the Yellow Pages. Different people will look for a business in different ways.
The problem of people finding your site becomes a third level, of equal importance to the other two levels, because of the unique nature of the Internet. Unlike the Yellow Pages of the brick-and-mortar world, alphabetical categories for a business just don't work very well. People can still find your site through branding, through frequent advertising, even occasionally through browsing. But most of the people have become much more accustomed to SEARCHING for what they want on the Web.
Let's say you wanted to buy a Harley Davidson on the Internet. How would you find one?
Chances are good you would go to your favorite search engine. If you enter "motorcycles" as your search term, however, you're going to be quickly inundated with several million irrelevant results. Even though that is the first level answer to our "What are you selling?" question, it's far too broad a keyword to get the buyer to where they want to go.
The second level answer is even more useless. If they enter "status" or "prestige" (which we know is what the customer really wants to buy) it's highly unlikely they'll ever find a site to buy their Harley.
The third level answer to our question is usually where the other two levels meet. They might enter "expensive motorcycle," or maybe "motorcycle sportster," or perhaps even "American motorcycle chopper hog."
NOW, we're talking keywords -- the words or phrases people actually use to search for something.
(Yea, I know the first search term they use will probably be " Harley." That is not where our first and second level answers merge, but rather is the result of very effective branding. It's important, but doesn't help you much unless the name of your business has been just as heavily branded. Someday soon, we'll talk about the importance of branding, but right now we'll continue to concentrate on the more short-term importance of keywords.)
----- Keywords are the Key -----
What are you selling? When you find the answers for all three levels to this question, you're well on your way to developing a keyword list. Just like goodwill, this is an intangible business asset.
Most people think of a keyword list in the same way a fourteen-year-old thinks of a chess game. Move the knight here, the bishop there, and it's a wham bam wanna-play-again game. You can play chess like that, and yea, you can put together a keyword list in maybe thirty minutes, but neither approach will make you a master.
Developing a good keyword list is a complicated, very iterative process that requires time and effort. But that's not a bad thing. If it was too easy, after all, your keyword list wouldn't be worth much as a business asset. The time and effort are investments that can eventually pay big dividends.
Unfortunately, to some extent, keywords fall into the chicken and egg paradox. Which comes first? You can't really fine-tune your keywords until you have an active, running business. But you shouldn't even think about starting a business until you have chosen your primary keywords. Conundrum city!
However, while you can't fine-tune or finalize your keywords you start using them, what you can do is understand some basic principles and concepts. With those as your foundation, there is hope you can select at least the most important keywords for your site.
----- Building a Foundation (or, This is IMPORTANT Stuff!) -----
If we were to graph the relative importance of keywords across the Web, they would be best described by a standard bell curve.
At one end of the spectrum are the keywords that are so highly competitive as to often be useless to you. Go to Google and search for "travel." Notice all the sponsored links? There might be a few government or educational sites (.gov or .edu), but you'll probably notice that most of the results are extremely well known, very long established web sites. Unless you have a few million dollars in the bank, I wouldn't recommended trying to compete with Microsoft or Travelocity for this particular keyword.
This side of our bell curve is filled by the words and phrases that are the most searched on the Internet. They include travel, mp3, jobs, sex (of course), music, food and many more. Millions of people enter these words into a search engine every single day. And thousands of web sites are trying to capture that very big audience. It's a bit like the lottery. The rewards for winning seem gargantuan, but your chances are abysmally slim. You compete for these keywords at your own risk!
Later, I'll offer some tools and references to help you recognize the highly competitive keywords. Much later, on another day, we'll also talk about ways you CAN learn to compete with the big boys with a margin of success. After all, you can't win the lottery unless you at least buy a ticket.
(Did anyone notice I switched from selling motorcycles to selling something in the travel industry? The best keyword for our hypothetical timepiece site would still include "harley," and I wanted to avoid complicating this discussion with trademarks.)
At the other end of our bell curve are keywords that are so highly targeted that John Q. Public will never think to use them when searching. The keyword "peregrination" may be a delightful synonym for travel, but it won't bring you a lot of visitors.
In the middle of our bell curve lies the bull's eye. These are the words and phrases people actually will USE when searching for products or services. And that brings us to another, albeit slightly different, way of looking at the same thing.
----- Painting a Target on Your Customer -----
Contrary to popular opinion, you don't really want visitors. You want customers.
If you could build a web site that came up on the first page when someone searched for "britney spears," you might get a ton of visitors. But you'd be a bit daft to think you could sell many Harley Davidson bikes. The traffic you received would be very untargeted and your bandwidth costs would probably far exceed the occasional and random sale.
Selecting the keywords for which you will optimize your web site is how you prequalify your traffic. Say that three times, real slow. Let it sink in. It's important.
If you select keywords that target your traffic, your conversion rate (how many people visit versus how many people buy) should be fairly high. But, uh, not too high. If your keywords are TOO targeted, you're almost certainly missing a large chunk of your potential market. Gee, that sounds a bit like a curve, too, doesn't it?
It's not really a different curve, though - it's a subset of the same one we talked about previously. The subset includes only the keywords specific to your product or service. At one end are the keywords with the potential to bring hoards of largely untargeted visitors to your web site, at the other end are the words that will bring a very few highly targeted people, and in the middle are the ones that represent an equitable ROI (Return on Investment is something we'll discuss in just a minute).
Unless you're selling U.S. dollar bills for eighty cents each, you will never want to promote keywords too far to the left on the curve. This is a critical point to understand, so let's talk about it for a minute.
----- Using ROI to Select Your Keywords -----
Simplified, we can say Return On Investment = Profit / Cost * 100.
Let's say you get 1,000 visitors per day to your site from the keyword "e-commerce," and it costs you one U.S. penny for each one. That cost comes from several different places, but we don't need to worry about those yet. You just invested $10 USD. Unfortunately, these are very untargeted visitors and you manage to sell only one e-commerce widget, for $5 USD. Your Return on Investment (ROI) is -50%, resulting in a net loss of five bucks. Oops.
Generic keywords can bring lots of traffic, but the traffic is very poorly qualified. They don't stick around your site, and they rarely buy anything. They are window shopping, plain and simple. That doesn't mean you don't want them, of course, because they do have potential value down the road. It just means you don't want to spend all your resources attracting them and, because these terms are often highly competitive ones, they DO require a lot of resources.
Increasing your search engine rank for the keyword "e-commerce" won't help!
Getting 10,000 equally untargeted visitors every day only means you lose $50 instead of $5. When you're losing money, you can't make it up on volume. A negative ROI is always bad. You are too far to the left on the Keyword ROI curve and the only solution is to move to the right, to a more highly targeted keyword.
Generic keywords don't always go negative, of course. But they do invariably result in a lower ROI, sometimes too low for a business to survive.
Okay, so you make some changes to your web site copy, targeting the keyword "e-commerce widgets in Southern Michigan." You now get 10 visitors a day to your site, and it takes a week to sell one of those visitors an e-commerce widget. Those seventy visitors still cost a penny a piece, seventy cents, and you still made $5 on your single sale. Your net profit is $4.30, with a very healthy ROI of 614%.
If you can increase your SE rank for "e-commerce widgets in Southern Michigan," you should get more visitors and your ROI will likely remain the same. Ergo, more money in the cash register. However, if you already have the number one rank for that keyword (and chances are VERY good you would), you'll have to learn to live on $4.30 a week profit. You have a great ROI, but miserable cash flow. Oops.
Of course, what's happened here is that you optimized your site for a keyword too far to the right on our curve. It is highly targeted, with good converstion and a really impressive ROI, but there just isn't enough volume because very few people search that way.
So, you go back and optimize for the keyword "e-commerce widget." You get 500 visitors a day and sell two widgets. That's a net profit of $5 a day, or a ROI of 100%. Note that your ROI has dropped from a very high 614%, but you're making substantially more money. (We are grossly over-simplifying ROI by including only variable costs, but it should still serve to make the point.)
Selecting your keywords is the foundation upon which everything else in SEO must be built.
----- Search Engine Keywords -----
The secret to SEO is to not select a single keyword, or even a handful of keywords. You wants scores, hundreds, possibly even thousands. Instead of a few highly visible keywords that bring in a ton of untargeted traffic, you want a hundred less competitive keywords that TOGETHER can bring in almost as much very targeted traffic.
So, how do you choose the right keywords. The first step in selecting the keywords for your site is to make a whole lot of educated guesses.
Try to find your competitors using just a search engine. Keep good records of what keywords you try and what works best. Note the rank for each keyword for each competitor. Did they come up number one on the search? Or were they buried on page nine? A spreadsheet makes an excellent tool for this early analysis.
The only keywords that really matter are the ones that work, but you can also often get an idea of what your competitors "think" are good keywords. Most browsers will let you right-click on your competitor's page and do a "View Source." Look near the top of their source code for something like META NAME="keywords" and then examine the Contents field that immediately follows it. These are the keywords your competitor has chosen to target, and they might give you a few fresh ideas. If nothing else, you can use these Meta Keywords to fuel additional searches.
Once you've built a list of keywords, take it and set it aside.
Depending on your industry and your own history within that industry, what you used to find your competitors might be of dubious quality. There's a real good chance you know too much to be a "typical" customer.
It's equally true that your competitors may be similarly blinded by their own industry experience. You may be finding their sites only because they targeted the wrong keywords, the keywords on the far end of the curve that only insiders would typically use. If you make the same mistake, the only people who will ever find your site will be your competition!
Ask your friends, someone outside your industry, to do the same thing you already did. Have them find your competitors using a search engine, while maintaining good records of what works and what doesn't work. They should record not only what keywords were used, but the position and page of the relevant sites found. The more people you enlist, the better will be your initial keywords list.
When you compare your multiple lists (yours and those of your friends), and combine them where applicable, you'll have the nexus of a keyword list. This is NOT your keyword list, though. It belongs to your competitors. It's how your competitors can be found. If you use this list, you'll be on a par with your competition (assuming you use it well). That's cool, but we want a birdie.
----- Extending your Keyword List -----
Now that you have a starting point using people (yourself and friends), we're going to see if we can use some software tools to extend the list.
To see this software in action, so we can better understand what to expect, we'll first look at it using example keywords. Go to AltaVista ( http://www.altavista.com/ ) and do a keyword search for "poetry." When the first page of results are returned, look for the RELATED SEARCHES section to the right. This is an often unrealized and untapped gold mine!
When people go to AltaVista and search for a generic keyword like "poetry," they very often don't like the results. Too much forest and not enough trees. So they try another search, a refinement of their previous, in hopes of getting better results. Indeed, studies have shown that most searches refine their attempts two or three times before getting good results. AltaVista tracks those refinements and when it finds a refined search used by a lot of people, it offers that as a suggestion to others. So, we know that people who start out searching for "poetry" will also very often search for "love poetry" and "poetry contests."
Can you spell i-n-v-a-l-u-a-b-l-e?
We did this first with a simple example, because AltaVista won't offer suggestions unless your keyword is fairly generic. You should try it with each item in your keyword list, but don't be surprised (or discouraged) if many or even most don't return a suggestion list. That just means you already have nicely targeted keywords. Where you do get a suggestion list from AV, however, you should use those to extend your list.
If you think this kind of research is time-consuming, entering each of your keywords into AltaVista, I almost hate to tell you that you're just getting started.
AltaVista isn't the only search engine that includes a suggestion list, and you need to duplicate your efforts on a number of others. Here's a list of some that will help you extend your keyword list - but you should also realize that search engines change almost daily and this list will never be all-inclusive. Every time you use a search engine, check to see if it has been updated to include the Suggestion feature. Then use it!
Excite http://excite.com/
Teoma http://teoma.com/index.asp
Lycos http://www.lycos.com/
WebCrawler http://www.webcrawler.com/info.wbcrwl/
AllTheWeb http://www.bos2.alltheweb.com/
Getting' tired, yet? I hope not, because we still have a few more stops to make.
Some day soon in the Gazette, we're going to spend time talking about what many call the PPC (Pay Per Click) Search Engines, but we need to understand a little about them right now. There are a lot of PPC Engines, but the two largest are O*****re and Google. PPC simply means that when a visitor clicks on the link returned in a search, the web site (who has now become an advertiser) will pay the search engine a few pennies (or dollars) for the visit. How much they pay can be determined in several different ways, none of which matter to us right now.
What is important to us is that the PPC Engines want people to advertise their keywords.
To entice web sites to advertise, some are willing to tell us how many times a specific keyword has been used in their search engine. Virtually all of the PPC Engines will at least give us some good ideas on selecting keywords that are both popular and relevant. Additionally, there are a few tools available designed specifically for PPC bidding.
WordSpot, for example, offers a paid service where they will create a fairly comprehensive report for your specific keywords. A similar paid service can be found at WordTracker, along with a free trial. Google AdWords offers a tool called Keyword Suggestions that gives you a rough idea of popularity (and does offer excellent variations for your keywords), though it doesn't give a true numerical index of the keyword popularity. Like Google's tool, O*****re's will offer variations of your keywords, but it will also give a true numerical count of the number of times each keyword was entered by a human searcher. A relatively new tool, at Digital Point Solutions, brings the results of WordTracker and O*****re together for a convenient comparison.
Here are the links:
* http://www.KeywordDigger.com
* http://www.wordspot.com/
* http://www.wordtracker.com/
* https://adwords.google.com/select/main?cmd=KeywordSandbox
* http://inventory.o*****re.com/d/searchinventory/suggestion/
* http://www.digitalpoint.com/tools/suggestion/
----- Putting It All Together -----
You now have the genesis of a Keyword List. Conservatively, you should have several hundred keyword phrases, and several thousand isn't at all unusual. So far, we've been learning how to move knights and bishops around the board. Building a comprehensive list of relevant keywords is a lot of work, but it's still just work.
It's time now, to move into Mastery.
To refine your list, we should ideally look at the ROI for each Keyword you've gathered. But, uh, what if you don't HAVE any returns on your investment, yet? That's what makes this a chicken and egg paradox. Your best keywords will depend on how much profit they bring, we can't determine profit before we promote the web site, but we need keywords before we start promoting. A vicious circle, indeed.
Remember our bell curve? One side is occupied by keywords that bring a lot of traffic but not necessarily a corresponding number of sales. Most of the people on the Internet make the mistake of optimizing for one of these keywords, often making those terms very competitive. I mean, only so many sites can get a first page ranking, right? You want to avoid these keywords, at least for right now. Let someone else fight over those. Your investment will be too high, your return too low.
The other side of the bell curve is occupied by keywords that don't bring a ton of visitors, but which otherwise convert really well. These are a real goldmine, but often represent a whole lot of work (we'll talk about why in a minute).
In the middle of the curve are the keywords that bring a decent amount of traffic that results in a decent amount of sales.
Your job, now, is to go through your list of keywords and assign each of them a priority number. Make the most competitive, highest traffic phrase number one. Then find the second most competitive, highest traffic keyword. Keep going until every keyword in your list is assigned a number, then sort the list accordingly. Obviously, there's going to a whole lot of guessing. How well you guess, in large part, defines your mastery of SEO. That's the part no one can teach you.
Ready? Go to the last keyword on your list, the least competitive, lowest traffic one, and type it into Google. Visit each of the ten sites on the first page and see if you can determine why the search engine though that site was relevant to that keyword. Ask yourself, could you build a page that was more relevant? If you can honestly answer yes, if you are confident you could get a web page listed in the top ten, go to the next keyword in your list and repeat the whole process. You want to work your way up your list of keywords until you are no longer certain of success.
THAT is where you want to start optimizing your web site.
You've just found the point on your bell curve that takes into account relevance, an estimate of ROI based on competitiveness, and your own gut feelings for your current ability to get results.
Create a web page optimized for that specific keyword. Get some links to it, both internal and, if possible, external. Then, while you are waiting for it to be visited and indexed, move down your list to the next keyword you feel confident is attainable. Repeat this process until you run out of keywords.
Of course, in truth, the process is going to depend on the type of web site you're building. You can't just build pages randomly, and the structure of the site will determine the best way to proceed. However, within that structure, you need to always keep your real goal in mind.
The definitive secret to SEO is deceptively simple, just like learning to play chess. Create a bell curve, pick a point on the curve, then optimize everything to one side of that point. In all likelihood, there won't be a single keyword on that side of the curve that will make your SEO campaign a huge success. Taken collectively, however, they're going to be dynamite. Instead of getting a lot of traffic from just a few high profile keywords, you're going to be getting a trickle from bunches and bunches of keywords, each of which will bring you highly targeted visitors. Make sense?
Here's the good part.
By the time you've exhausted the keywords on one side of your originating point, you should have a really solid foundation. You've built a ton of great content, which in turn, should have brought you some good links. It's now time to return to that point on the curve and start building web pages for the keywords going in the other direction. With your new foundation, you're going to find it much easier to start attaining some of those more competitive terms.
The secret to SEO is keywords, and the secret to keywords is targeted diversity.
* Epilogue (or, Do Not Open Until Christmas) : As your SEO campaign progresses over time, you want to periodically revisit your keyword list and convert all those guessed ROI numbers into real ones. Use your server logs to track which keywords are bringing in visitors and, where possible, which visitors are converting to customers. The more you know about your keywords, the more success you can expect.
Highly Competitive Keywords
In our lead article this issue, we briefly talked about highly competitive keywords. These are the search terms that are among the most frequently used on the Internet, words like "travel" and "music." Most of the time you'll want to avoid targeting these keywords. Competition for these keywords is fierce and going head-to-head with the Big Boys is often a mistake. It will be hard to get a decent ROI because your Investment will necessarily be high to compete.
To avoid them, obviously you need to know what they are. And that's my excuse for providing this list of resources where you can find the most common search terms used across the Internet. Beyond excuses, however, it can be just plain fun to discover what interests the world this week.
Wordspot ( http://www.wordspot.com/ ) is my personal favorite. They used to provide a list of the top 200 keywords directly on their site, but have recently changed their policy and now give it only when you sign up for their newsletter. But that's okay. With the newsletter coming directly to your email box, you won't have to remember to occasionally check on their list. As with all things on the Internet, their list of words should be taken with a grain of salt. Their methodology is good, always getting better, but certainly not perfect.
CNet ( http://www.search.com/top?tag=ex.se.sr.top.top2 ) offers a list of the top 100 keywords from their own search.com utility which should again (need I say it?) be taken with a grain of salt. While this list is derived from a very substantial pool of data, it doesn't necessarily represent the Internet's "typical" searcher. CNet, after all, tends to attract the more computer-savvy user.
The Lycos 50 ( http://50.lycos.com/ ) is updated weekly to reflect the top 50 searches at Lycos. I particularly like the way they also show how many weeks a keyword has made the list, as this is obviously important information. If your product or service is seasonal, be sure to check out their archives, too.
Yahoo Buzz Index ( http://buzz.yahoo.com/ ) is very similar to the Lycos 50, but is limited to the top 20 keywords from the extensive Yahoo database.
Google Zeitgeist ( http://www.google.com/press/zeitgeist.html ) pulls from an absolutely huge database of search queries, is extremely well organized (especially if you're interested in non-English keywords), but is sadly limited to just the Top 10 keywords in multiple categories.
By Ron Carnell
Comment