9/30/2008

Notes on keywords, niches, and real life.

Just some notes and updates on keywords, niches, and how it really all fits together.

Been going through several trainings lately. They all have their points.

But they each have some confusion on where keywords work and why. I've earlier pointed out about some keywords are "natural". Means keywords that people use are simply ordinary to them. That turns out to be the basis of all search - find out what people are talking about and the words they use to describe them. When a lot of people use the same terms consistently, these have been named "keywords".

The trick is that these also tell a lot about where peoples' attention is at - and where they continue to look for things in certain areas, these are known as niches. Some have more activity than others - and this is known as "competition".

Now, the way search engines have looked for keywords and evaluated sites has changed dramatically over the years. This is because there has been some artificial importance to the top few sites which search engines turn up. Because Google shows up only a certain few web locations for each particular set of words. And some have gone for the short-term profit of trying to grab and hold onto these "top sites".

Meanwhile, a whole 'nother section of the web developed, called social media. These are what was formerly known as "communities" and they still are - it's just that social aspects have been high-lighted with new technology and people have de-centralized and more loosely-connected themselves.

While search engines have always simply tried to group and serve up what is really important to the Internet - they are now looking to the social media to point their way. This is seemingly a very good rating system which is almost impossible to spam - so you get routinely great content. So these are now what the search engines serve up as tops.

This comes back to keywords, because - again - real keyword use is natural and organic to the traffic that uses them.

Now, where this goes off the rails is people thinking that keywords and niches are the same things. True, niches have words and a keyword phrase might also be a niche.

But there are differences - niches really can't be "dominated". They are constantly moving targets. (I really dislike that word - almost as bad a "manifesto" - they both have some negative emotive content...)

Ok, so you can hit that moving target here and there, but you can't ride it like a bull, even for 8 seconds. It just moves too fast.

Buuut - you can keep pages in top search engine positions for "long-tail" keywords. Those are like Chris Anderson's "Long Tail" of marketing - where very precise mini-niches have almost vertical markets and change very little.

Jack Humphrey has built most of his work on this exact point - becoming an authority in these niches and then monetizing from that vantage.

You essentially create a volume of work which is all great content and develop trust with all these niche denizens. And they start looking to you as a dependable resource and trust your data. As you do, they will also click through on what you offer. And that's where you get your funding - because you either then provide them good content or get paid a commission for having a link on your site where people who trust you will click through and by their products.

OK?

Now, here is the simple point that people get waaay wrong: Niches aren't keywords. No, duh?

Niches have lots of competing pages for the same basic terms. Because a lot of people are talking about the same stuff and search engines have difficulty discerning which one is more important than others. Search engines are basically linear.

In social media, you don't have that problem, because all your content is referred by your "friends" (people you trust as authorities) and you don't really have to search for much. And meanwhile, most of these social media sites acquire some sort of search capacity as they grow. Primitive search, but it's useful.

But search engines are necessary part of life right now. They allow you to bridge between niches. You can keep track of related niches this way - or find out about something which is of interest to you, but isn't perhaps all that popular in the niches you are interested in.

So - in marketing (working at getting paid for offering solutions to people you don't particularly know) we want some tools to enable us to get our pages up on search engines and keep your pages up in the prime categories and so people can find your stuff - and then pay you for your work.

All long-tail keywords aren't the same.

Niches are high volume traffic. Plenty of demand, plenty of supply.

Long-tail niches have much less demand, much more focused supply.

What you are after is to optimize your pages for the search engines - so that they will see your pages and say they are important.

Right now (and things are constantly changing) all you need are
  1. you keyword phrase in your page title
  2. your keyword phrase in your headings
  3. that keyword phrase in your incoming links
  4. and the data on your page defines what you are talking about - your topic theme. (If you want more data on this last, look up LSI - Latent Semantic Indexing)
Other than that, you want fresh original content - which the social media discover and talk about. (I could go on at length about inspiration and how to improve on this - but see my other sites and blogs for this, as least for now...)

Here's the main point I wanted to talk to you about today:

How do you figure out what keywords to use to put in those headers and so forth?

You can't go by any "magic number" per se. There isn't a certain number which is big enough to go after, or competition which is small enough.

You can look in Google for certain keywords and find how many pages have that keyword phrase (put the entire phrase in "quotes") compared to how many pages have all those words - or most of them - on that page somewhere. This gives you a rough idea of your actual competition, different as the wheat from the chaff that surrounds it while it grows. You are harvesting here.

And if you have those four points in above, this will usually do the trick.

But what about those main niches - you'd think there would be more marketing possible when there are a lot more traffic?!

True enough. The trick is in what we've already covered: lots of disoriented traffic. Search engines have a hard time with their millions of pages in serving up the right ones.

Here's two tools you can use:

1. Go small first, and then build your foundation so that you can reach the sky. And it's about that much work as well. Grab a sub-niche by providing good content and optimizing it for search engines (called SEO, btw). Then grab the niche next to that. [Ex: border collie training videos, miniature collie training videos, miniature collie training DVDs, border collie training DVDs, etc.]

That's going to take some regular work, but practically, it is one of the best ways to start.

2. Find areas where there is high demand, low supply. One author (Sumantra Roy - http://www.1stSearchRanking.com) figured out a formula for comparing the numbers of requests for data against the number of actual pages across different search engines (they all use different algorithms and get different numbers of pages for the same terms). He called this "KEI" or Keyword Efficiency Index (or something like that - I'm writing faster than I can research right now...)

Some paid services such as Wordtracker are built specifically on this one feature - although they've expanded their offerings as they've been successful.

And this was what through me off this area for a long time, since I didn't know that a person could "roll your own" KEI. You can do this if you get a tool which will tell you the amount searches and the amount of pages on various search engines. One tool I found to do this is called "Niche Keyword Buzz". (And this company makes a lot of other free and useful little tools as well...)

This tool actually uses the free Wordtracker keyword service and gets the words you are interested in, then goes to Google, Yahoo, and MSN to tell you what the competition is.

You probably already know about Google's Adword KeywordExternalTool - which can give you a set of words which are related to the words you are searching for. And you can save that list of words as a text list.

Take that keyword list and plug it into Niche Keyword Buzz (or similar tool) and then run the results. Now, export this to a CSV and open it in OpenOffice (or Excel) and then add a column which does the calculations I laid out above.

Once you get the results (you may need to multiply that result by 1000 to get it out of decimal-land), then you can see that certain of these long-tail keywords have high demand, but relatively few pages. (Obviously, this needs some work - and your mileage may vary...)

But that's the point. You spend your time on niches that have a lot of traffic, but are under-served. You don't spend hours on content that gets buried by the search engines - or nobody is looking for. That is your main key to success in this area. Figure out your own KEI or hire WordTracker to do it for you. And there are other services out there.

And there you have it - my two cents. I'll keep working on this as I can and let you know, but I haven't posted here for awhile and thought you deserved it. ;)

- - - -
update (some hours later)

Above tool isn't heavy duty. It times out and has connection errors if you push it too hard. Nice start, though.

And found this definition for KEI:
The Keyword Effectiveness Index (KEI) is the square of the popularity of a keyword multiplied by 1000 and divided by the number of sites which appear in search engines for that keyword. It is invented to measure which keywords are worth optimizing your site for and used by Search Engine Optimizers (SEO). The higher the KEI, better the keyword.
In this case we took the Overture popularity count of a keyword and the number of sites which appear in Google for that keyword. Other counts and other search engines will give another outcome.

Note that Overture isn't even able to be accessed these days (and was nearly a couple years out of date), so the above tool on that site returns nothing every time. WordTracker and others have developed workarounds.

Wish I could have one site with all this data, rather than having to scrape it...

- - - -

Further, a discussion about using WordTracker from http://www.buildwebsite4u.com/building/keywords.shtml
"Write down the most profitable keywords (KEI > 10) along with their attributes, click the link at the bottom of the page to try again, clear the basket and type in your next keyword. You may repeat it as many times as you need to analyze all related keywords.

Note:

  • WordTracker assigns high KEI to words with low Count and very low (or zero) Competing. Don't target these keywords. What good would it do for you to rank high for these words if no one ever searches for them? Select words with high Count (not less than 100)."
- - - -
And a discussion of KEI and its pitfalls is here.

- - - -
Found a free (47Mb download) tool called Rank Checker, which seems like it's set to do the job I need (and then some). Still evaluating - get it at http://www.link-assistant.com

Now a point I may not have made clear in the above:

You are looking for popular phrases. Simply. But you don't start and end with good KEI keywords. It's probably more the start of what you do, then more research, and then the end of that research.

You want to get a smart niche where there is high demand, low supply. But then you check into that niche fully to make sure you can generate content for it and that you can actually monetize it and get something from it. Lots of traffic, but no buyers - well, that's a poorly-run non-profit site that can't even keep itself afloat.

Once you know your niche is something that you're interested in, something you can generate tons of content for, and an area which has buyers and does exchange money - at that point you can then start finding the keywords you need to use to generate that content.

Now, here's the interesting point: as you set up these keywords, and optimize each page, you'll find yourself ranking for other keywords that you didn't even particularly target. Remember that "in quotes" tool above? That tells you what your actual competition is. But as search engines aren't all that specific, they are also awarding you top spots for keywords you didn't really target. So, as long as you keep optimizing your pages you will actually take on the competition in areas you weren't even targeting.

Like "dog training". Too much competition. But - you can make lots of contents for "training chiuhuahua dogs", "training boxer dogs", "training collie dogs", etc. As you keep building up all these sets of pages, videos, and podcasts for longer-tail keyword phrases, you site actually moves you up the rankings on "dog training" and also "dog" and "training".

Nice to know.

- - - -
Update: 1/10
This from adyacker.com:
Try to choose words with a KEI score above 3 whenever possible, and also make sure that term has a decent amount of daily searches. Even if the term only has 400 daily searches, you’ll be able to drive much more traffic with a KEI score of 3, than a phrase with 1000 searches and a KEI score of .18.
Now I took another version of the "Buzz" tool above (called Matt's Free Keyword Tool) and used it to query wordtracker and get the data. It got down to about 88 keywords out of 100 before it "lost connection" (meaning that tool was done for the day as Google and the other SE's had called it quits for that logon - so I can run this once every 24 hours?).

I cut out all the data which was missing slots (like nothing from Google, for ex.) and save the data to CSV format.

Opening this up in Open Office, I then entered the KEI formula above into an empty column and got some fascinating results. You just don't see this stuff by looking casually at the numbers of competition and searches. And you can't get this data unless you have a program which will query the searches and match them up with the competition - and then you can calculate their actual worth.

My problem was being stuck in a keyword "Self Help" which I knew to have lots of traffic. But variations of this didn't get me anywhere, since most programs only give you another word tacked on. The KEI of these variations got me nowhere. Now, when I started using various synonyms for the benefits people got out of self help, I started getting amazing results.

Look up health, success, wealth, happiness, achievement - just for starters. Lots of traffic, tons of long-tail niches, many with decent daily searches, but few competing pages.

I took success for one such search through the "Buzz/Matt's" tool and got 33 usable high-traffic (30 or better daily hits) and a KEI over 3. Some were fairly tight niches, where off-hand I could only write maybe one article without some research - but when I see that "key success factors for hotel web sites" has 119 searches daily and a KEI of over 54,000 - you know I could set some time aside for that!

Now, I'd then take these hot KEI-proven keywords and then run them in some sort of competition checker, or just look them up on Google and use my SEO Firefox plug-in to see the data right off. (all in title, all in link, etc.) And I could then study up these guys to see what they are talking about and how they are presenting themselves to see how I would set up my pages and presentation to out-create them.

Social media just with those keywords would take over top spots, but I'd also then set the back up pages so that there is something for them to post to.

Previously, I said to make mini-webs and have your social media posts point to them. Here's the new "social media full court press":
  1. Write articles - six versions of the same thing - with the same keywords. Post the best one to your blog and the others to the top 5 article directories.
  2. Podcast the best version and post it to archive.org
  3. Create a powerpoint and post to Slideshare.com
  4. Link the podcast to the powerpoint and create a slidecast on Slideshare.com
  5. Create a video from this same data and submit it via TubeMogul.com
  6. Create a Squidoo lens when these articles go live and link everything through that lens.
  7. Social post everything through OnlyWire as you create it (don't try to post all of these the same day - you could get your account pulled for spamming.)
  8. Everything points to your original blog post - but now create a mini-blog summary and put it on Blogger and/or other free blog.
  9. Update your original blog post with links to this other data.
And sure, this is a lot of work. But you are getting most of the above-the-fold positions in Google. So you should also be able to monetize this traffic meanwhile.

Now, back to Campbell, who got me started on this (and I only put it aside until I solved a non-Wordtracker solution, as outlined above) - he said to then find an affiliate product which you can sell which matches those keywords. So you then create the sales page first and all this promo would run straight to that sales page - if you are only doing "bum marketing".

But... his later stuff says to have this sales page simply linked to your blog page - in a key position so people can find it easily. (Top left...) And meanwhile, create an autoresponder series which supports this keyword and also a giveaway (ethical bribe) so people will opt in for your data.

That's the long and short way of monetizing right off the bat. And since we are in this to survive, you have to have some sort of way to monetize as you go. Every link on a page needs to monetize in some fashion - either as a (cloaked) affiliate link or to a direct product you sell. (Or simply have a donate page.)

Putting these links on a blog page builds your trust with them until they are willing to clickthrough on one of your links (or simply give you money in appreciation).

- - - -

Now, let's review our keyword data:

  1. Paid route is through WordTracker to find popular keywords which have low competition.
  2. You want niches which have decent traffic and but are poorly served, or the pages which do serve them are poorly optimized (most are). KEY POINT and PURPOSE.
  3. Unpaid route is getting a tool like Rank Tracker to find your keywords and do the KEI on them. Alternates are the "Buzz/Matt's" tools above, and then doing your own KEI through a spreadsheet (if you have lots of time spread over short periods daily). (But I would recommend you download one of these and try out what I just outlined above to get your hands really dirty in what makes KEI so important.)
  4. Get your niches and then do your full research on these to see what the supply has been offering and how you can create better stuff.
  5. Find some products you can sell to that niche.
  6. Pick the most likely niche and start marketing to it via social media submissions (this assumes you have a WordPress or similar blog on your own site - so they can arrive there).
  7. Set up your backend to be able to deliver to this traffic as it starts converting.
  8. Meanwhile, keep promoting to that niche until you have several blog posts and lots of videos, etc. about that particular keyword phrase - essentially when you've taken the bulk of the top search engine spots.
  9. Then take your next niche keyword phrase - which is another variation of the shorter niche keyword phrase you want to really want to take over.
  10. As you start controlling more and more of these longer phrases, you will start showing up for the shorter phrase.
That 10 point step (although the research phase is many steps all in itself) really summarizes and boils down the data I've been studying from this over-paid and free training I've been taking lately.

One more lesson tomorrow and then I'm done with all this and straight onto real work.

Once I do get thoroughly on the way to my Online Millions, then I'll let you take my course.

- - - -
Update: 10/2/08

Cute trick with that KEI tool - Rank Tracker. Unless you buy their pro version ($98) you can't save your work. And you can't copy/paste the results. So you're going to have to hand copy your data. And start over every time. Cute. Wish I had a thumbs down right now... But the program is great otherwise (and it's Java-inherent slowness).

No comments: