Effective Keyword Discovery and Traffic-on-Investment

Keyword Discovery Matters More than Ever

As search engines mature, keyword discovery becomes more important. The competition within your industry for commonly known terms will increase and, more importantly, your customers will become more savvy and using search engines in ways that find the correct content. No longer will ranking for the generic term “baseball” work, if you are selling vintage Topps baseball cards. For Search Engine Optimizers, this becomes even more important. Good Keyword Discovery means happier clients – they will receive a far quicker return on their SEO investment, you can manage their expectations by knowing more accurately the number of links required and expense involved, and can accomplish better budgeting overall. Finally, and most importantly, good Keyword Discovery means higher profits: using the least amount of money and time to accomplish the most in the search engines.

It is actually a really simple equation, you can probably figure out a much better one!
a simplified version of a keyword efficacy equation

What is Traffic on Investment

First, let’s get something clear. Keyword Popularity is Meaningless without Competitive Data and Keyword Competitiveness is Meaningless without Competitive Data. Individually, these pieces of information will poorly inform your strategies and often result in over or under-spending – lowering profits and overall Traffic on Investment. Predictive ToI is, simply, Popularity divided by Competitiveness.

Keyword Sources

While traffic data is easy to come by, you must concern yourself with the integrity of that data. There are a couple of questions that need to be asked about your source – whether that be Google’s suggestions, Yahoo, KeywordDiscovery, Wordze, Wordtracker or others.

  1. Does the Data include Automated Searches?
    Considering the prolific nature of rank-checking software, popular keywords can be greatly overestimated. If a keyword had 100 unique competitors, and all of them are checking rankings daily, you are generating at least 3000 extra visitors every month, unless the keyword data service filters these out.
  2. Is there an incentive to game the data collection?
    Yahoo’s keyword suggestion used to suffer heavily from spammers spamming spammers. Essentially, spammers would generate false searches so that their websites would show up in top keyword lists, making it likely that auto-scraping software would generate content based on their sites.
  3. Are there seasonal adjustments?
    I prefer keyword tools that give you the average over a larger period of time – at least 90 days. If you are only looking at a months worth of data, you could have highly skewed numbers.

One important item to note is that it is not necessary for you to have the exact number of searches to successfully determine Traffic on Investment. While it would be a best case scenario to have the exact number, it only truly matters that the data is internally relative and consistent. If the number of actual searches for baseball is 1000 but your data set says only 100, that is not necessarily bad. If the actual number for baseball card searches is 300 but your data set says 30, then your Traffic on Investment calculations will actually be identical in relation to one another. Traffic on Investment is a ratio and is only relevant when considered against other keywords. Subsequently, if you use the same datasource for your popularity number, and those numbers are related to actual results in a similar manner, then you will see little to no quality loss in your final measurement.

Keyword Competitiveness Variables

When considering the various measurements for keyword competitiveness, you need to be able to judge the actual value of each data point so that you can weight them by class. There are two questions that I ask about each measurement before I consider including them in any sort of ToI algorithm. I will apply these questions the factors I list below.

  1. To what degree does this measurement indicate a webmaster’s intent to compete for the targeted keyword?
  2. To what degree does this measurement indicate an obstacle to ranking for the targeted keyword?

Keyword Competitiveness Factors

Below is a list of common keyword competitiveness factors, some of which I include in the Keyata algorithm.

The Aggregate Variables

  1. Total matches (t)
    total number of results returned when the keyword is searched without quotes
    shortcomings: large volume of incidental occurrences of keywords.
  2. Exact matches (e)
    total number of results returned when the keyword is searched within quotes
    shortcomings:some incidental occurrences of keywords.
  3. InURL matches (u)
    total number of results returned when the keyword is searched within quotes after inurl:
    shortcomings: overstates competitiveness of single keyword phrases
  4. InAnchor matches (a)
    total number of results returned when the keyword is searched within quotes after inanchor:
    shortcomings: very few, may over accentuate competitiveness of brand names
  5. InTitle matches (l)
    total number of results returned when the keyword is searched within quotes after intitle:
    shortcomings: some incidental occurrences, especially of single keyword phrases
  6. InText matches (x)
    total number of results returned when the keyword is searched within quotes after intext:
    shortcomings: some incidental occurrences of keywords

The Ranking Page Variables

  1. Ranking Page Links (pl)
    total number of inbound links to the page ranking for a particular keyword
    shortcomings: many of these links may not include targeted keyword.
  2. Ranking Page PageRank (pp)
    Google PageRank of page ranking for a particular keyword
    shortcomings: slow updates.
  3. Ranking Page InTitle (pt)
    occurrence of keyword in title of ranking page
  4. Ranking Page InText, H1s, etc. (px)
    occurrence of keyword in various HTML tags of ranking page
    shortcomings: can be incidental. obstacle easy to overcome

The Domain Variables

  1. Ranking Domain Links (rl)
    total number of inbound links to the domain ranking for a particular keyword
    shortcomings: many of these links may not include targeted keyword.
  2. Ranking Domain PageRank (rp)
    Google PageRank of domain ranking for a particular keyword
    shortcomings: slow updates.
  3. Ranking Domain Age (ra)
    age of domain ranking for a particular keyword
    shortcomings: index age of domain or actual age? impacts certain engines differently?

Common Misleading Variables

  1. Bid Values
    average bids for a particular keyword in a paid-search listing shortcomings: does not indicate interest in competing via organic listings, does not indicate an obstacle for ranking organically, data is greatly skewed by paid-search policies (makes pharma look very uncompetitive), data is greatly skewed by price of item for sale.
  2. Traffic Estimates
    estimated traffic for a particular domain or page shortcomings: does not indicate interest in competing via organic listings, does not indicate an obstacle for ranking organically, data is skewed by type-in-traffic, paid-search campaigns, viral nature of site, etc.

Multipliers and Mathematical Fixes

What Matters the Most

Here is where, as an experienced SEO, you get to work your magic – make your own algorithm. Think critically about all of the individual factors and their shortcomings and begin to create multipliers for each of the items. Perhaps you think that the intitle number is very important, so you will give it a high multiplier in your equation. Perhaps you think that the incident rate of inurl is somehow dependent upon inanchor, so you create a multiplier that is actually an equation itself to depreciate high inanchor occurrences when there is also a similarly high inurl occurrence. All of these go into creating a final algorithm for keyword competitiveness.

Essentially, you will use math to devalue the less important, less meaningful variables and to empahsize those variables that mean the most. For examples, I always show great emphasis towards the allinanchor: aggregate data and the inbound link numbers to bot the domain and the ranking page (the link is KING). More importantly, links represent a legitimate obstacle to ranking – it is much easier for me to add a keyword to the title of my page than it is for me to secure a quality inbound link.

It is actually a really simple equation, you can probably figure out a much better one!
a simplified version of a keyword efficacy equation

No tags for this post.

3 Comments

  1. SEO'ed
    Oct 27, 2008

    Interesting equation. Though I got to admit, my approach is much simpler 🙂

  2. Will
    Oct 29, 2008

    That seems horribly complex. I’m sure your approach is thorough but I’d much rather have somebody do it for me. I personally prefer http://www.seokeywordranking.com, but I know there are others out there.

    Author’s Note: It is complex for a non web developer. To be honest, the solution start to finish took about 8 minutes for me to implement. But, I do have a servers readily available and have been doing this a while, so it made sense from the start.
  3. Carl
    Nov 6, 2008

    Good article. This is exactly the way to do it. It is not complex, just a weighted average. I have used a similar system for my keyword research process. The difficulty lies in the adjustment of the parameters. Could be simplified further by removing the least important factors though.

Trackbacks/Pingbacks

  1. A Ghoulish Collection of Links | This Month In SEO - 9/08 | TheVanBlog | Van SEO Design - [...] Effective Keyword Discovery and Traffic-on-Investment [...]
  2. Link Love - Monday 03/11/2008 - [...] Effective Keyword Discovery and Traffic-on-Investment on TheGoogleCache. They had covered many things about the keyword research and traffic in…

Submit a Comment

Your email address will not be published. Required fields are marked *