Printer Friendly Version
Email this thread to a friend
|
Featured Web Site Template |
|
There are 0 guests and 2 members in the forums right now.
Reflects user activity within the last 5 minutes
|
|
| Member |
Message |
JonWittwer
Joined: Eons Ago
# Posts: 32
|
Posted: 08/08/2003 11:51 am
Hi all,
I have been doing a lot of experimenting lately with all kinds of different formulas for evaluating keywords. I'm sure most of you are aware of KEI (Keyword Effectiveness Index).
MY goal is to develop a KPI (Keyword Profitability Index). There may already be something out there, but I haven't seen it, yet.
Here are some of the variables that can be used for the equation:
a) frequency -- obtained from Overture's keyword suggestion tool, from your PPC stats (impressions), or from Wordtracker, etc.
b) keyword variations -- the number of related keywords obtained from overture's suggestion tool, or simply the number of
c) total frequency -- obtained when adding up the frequencies or impressions for a set of similar keywords (i.e. for an entire Google AdWords campaign or Overture category).
d) number of PPC bids -- helps determine competitiveness.
e) maximum PPC bid -- this one is obvious
f) slope of PPC curve -- this one is getting complicated, but it could be a useful metric when you know the maximum, minimum, and the number of bids.
g) number of web pages -- the number used in the KEI formula for example
h) number of merchants -- for e-store owners, the number of stores in Yahoo or eBay associated with your keywords.
i) number of products -- associated with the merchants - the number of items from eBay or products in Yahoo, or where-ever. This variable is different than merchants because some merchants sell tons of products, and others focus on a few niche products.
If you want to list more variables, give them new letters ( but don't use "n" or "o" ) so that they can be discussed.
So, any suggestions on formulas that you have seen or used? The point is to be able to identify the ideal keywords to focus on, BEFORE you know click-through rates or conversion rates. In other words - an equation that could be used in connection with KEI to help zero in on good keywords.
To get you started, here is an equation that I have found to work in a few cases: (a/b^n)/e for (n = 2)
And, of course, KEI = (a^n)/g for (n > 1)
Many of you might not want to give away your secrets, but I guess you can either help out, or wait until I figure it out, myself.
[ Message was edited by: JonWittwer 08/08/2003 11:50 am ]
|
 |
yellowwing
Joined: May 21, 2002
# Posts: 2526
|
Posted: 08/08/2003 02:04 pm
I have always asked the merchant themselves. What is your most profitable product/service? I then look up the relevant keywords and make a plan to promote them.
I asked a new pharmacy client about a very popular pharmaceutical. He said that he only made $3 per prescription on it. "But if you can promote this other one, I'll make $50 per perscription!" So we did.
Openly communicating with the client is not really a secret, it's just good business.
|
|
Ron Carnell
Joined: May 15, 2001
# Posts: 206
|
Posted: 08/08/2003 10:16 pm
I've always felt KEI was fundamentally flawed, and would caution making against the same mistake.
To measure your competition, you first have to define it. Anyone here really think that guy on page 762 is going to take one of your clicks away from you? If you can even spell SEO, you're going to do better than that. He isn't any competition, so why should he be counted? It may not be quite so obvious, but if your goal is to hit the top 30, that site at #31 isn't your competition either.
The number of sites returned for your keyword means very little. Your competition will be defined by the strength of the sites you need to beat, not by the number.
|
|
JonWittwer
Joined: Eons Ago
# Posts: 32
|
Posted: 08/08/2003 11:28 pm
I would agree that the total number of pages is not a realistic indicator. And Ron, you make a good point that you have to know the strength of the sites. Perhaps the sum-total of the PageRank for each of the top 10 listings at Google, or some other measure of link popularity for the top 10 listings, would be better than simply the number of pages returned.
So, here we have another couple other variables -
j) competitive strength -- the sum of the PageRanks for the top 10 Google Listings.
k) collective link popularity -- total or average link popularity of the top 10 search listings.
One point of this exercise is to try to think of ways of converting somewhat intangible characteristics into meaningful numbers. After gathering all these possible variables, we can start having fun with equations and interpretation. Thank you for your contribution, Ron.
|
|
excell
Staff
Joined: Mar 19, 2001
# Posts: 14504
|
Posted: 08/09/2003 01:38 am
There are other strengths a site might have besides link popularity & PR.
One site may be strong in theme and run multiple pages across very close but slightly different key phrases or alternative phrases. It could have 10, 100 or a 1,000 pages ready to come into play, yet it is only one site (one of your competitors).
Whereas formulas and rules of thumb can be helpful in standard situations, I take the approach that no situation is standard and needs to be studied individually.
Your thought track is interesting and your discovers may be even more so, so keep going!
By the way, why aren't we allowed to use letters n) & o) 
|
|
Ron Carnell
Joined: May 15, 2001
# Posts: 206
|
Posted: 08/09/2003 06:15 am
I agree with excell. PR is only one indication of competitor strength, and for most keywords it would be a very weak indicator. On-page KW density would be a much better indicator, but ultimately the only truly accurate measurement would be a duplication of Google's algorithms. I suspect a lot of trial and error could get you close to that goal (and it would be fun to try), but it would never be absolutely accurate because Google has off-page data you can't readily get. Still, for most keywords, I suspect "close" would good enough.
Such a formula would benefit the more inexperienced SEO, but would probably only confirm what a more experienced SEO already knows. Ten minutes on each of the top ten sites, and a good SEO will have a pretty good idea how long it'll take him to be able to break into the list.
|
|
JonWittwer
Joined: Eons Ago
# Posts: 32
|
Posted: 08/09/2003 10:10 am
Let me mention the goal that I have in mind. When a person already has a business or is in the planning stages, it makes sense to consult the expert SEO, but what about the people who don't know what they want to get into and are scouring the internet for some kind of niche they can fill?
So let's move away from what we don't know or data we can't get. Let's assume that I AM Google or I AM Overture/Yahoo (or both). What could I do with the data I am collecting or already have that could let a person enter a keyword and find out how feasible it would be to enter the market based upon that keyword?
Forget about reality for a while. Tap into your imagination, and dream of what you WISH you could find out from a simple keyword search - then try to think of a way to convert the concept to a number. Some things WILL translate, other's may not. Some of the best ideas actually come from considering the most absurd solutions.
excell's "theme" is definitely something that would come into the picure. So a couple more measures of competitive strength...
m) keyword PPC spread -- the total collective strength of a competitor's Ad Campaign. How to quantify?
If I were Google or Overture, I would in fact have the capability of analyzing current PPC ad campaigns to determine how a company is grouping their keywords - and I would have statistics that indicated CTR, impressions, etc.
p) SEO effort -- this would be an indicator that a page is making ongoing efforts to maintain their top listing status. How to quantify?
Again, this one might not be feasible by anybody but the actual search engines themselves.
Why no n, o, l, etc.? Well, "n" is a good place-holder for an arbitrary number (as in the KEI). "o" and "l" can get confused with numbers.
|
|
JonWittwer
Joined: Eons Ago
# Posts: 32
|
Posted: 09/26/2003 12:16 am
This is the latest in my quest to find a good niche keyword formula.
The motivation for this particular post is that it is easy to get highly inflated KEI values for very general keywords ( like "internet marketing" )
This experimental modification of the KEI uses a special weighting factor to penalize overly popular keywords by taking into account the number of keyword variations (particularly when using Overture's search term suggestion tool)
To down-weight general (non-niche) keywords, you can divide the normal KEI value by:
b^(EXP(b/100)-1)
where:
"b" is the number of keyword variations returned using Overture's search term suggestion tool.
Here is how it works:
When b = 1 KEI is divided by 1
When b = 30 KEI is divided by 3.3
When b = 100 KEI is divided by about 2700
This weighting factor can be modified by changing the "100" in the equation.
[ Message was edited by: JonWittwer 09/26/2003 12:24 am ]
|
|
thejenn
Joined: Aug 08, 2001
# Posts: 9196
|
Posted: 09/30/2003 10:32 pm
Moving over to Pro SEO issues so that it can get some feedback from our more SEO focused members as opposed to our marketing focused members...
|
|
You are not permitted to post messages in this forum or topic, because of one or more of the following reasons:
- You have not yet logged in, or registered properly as a member
- You are a member, but no longer have posting rights.
- This is a private forum, for which you do not have permissions.
If you are a recent member, it's possible that you simply have not yet confirmed your account. Please
check your email for a message entitled 'JimWorld Forums: Confirm Your Account' and follow the instructions
contained within.
If you cannot find this message, click here to Re-Send it.
|
If you are still experiencing problem, please read the
Login Assistance
Article for some advice on what may be causing your login not to work properly.
|
Switch to Advanced Editor and ...
Create a New Topic
or Reply to this Thread
|
|
