|WordTracker vs Google AdWords|
Does anyone know how to correlate Google AdWord impression figures to the WordTracker numbers? In WordTracker there is an explanation of the ratio between the number of terms in their database (300 million over 60 days) and the number of daily search requests for all other search services (about 319 million/day). How can I get a reliable ratio (expressed as a percentage) to use for Google and WordTracker to be able to understand how much of the daily search each term is?
Does my question make any sense? :)
Welcome to WmW Shabba!
From the WordTracker website:
"Our main sources are Metacrawler and Dogpile, the two largest Metacrawlers on the net."
I am not sure that this is an accurate reflection of search traffic because different types of people use different engines for different purposes (wow, got all the variables in one sentence). But maybe I am off base here, could you please restate your sentence "How can I get a reliable ratio (expressed as a percentage) to use for Google and WordTracker to be able to understand how much of the daily search each term is?" Are you talking about useage, how many times a word is searched for?
Thanks for your reply!
I meant how many times a word is searched for.
Sure thing, glad to help.
Are you familiar with the goto.com search suggestion tool? If so, then skip this. If you are not familiar with it, click over to GOTO Search Suggestion Tool [inventory.goto.com] and type in your keyword. It will tell you how many times that keyword was searched for on goto.com last month, as well as other keyword combos to try, and how many times they were searched for.
As a reference for you, the number of total impressions on goto.com for my Adwords ad last month was about 20% higher than the total number of searches recorded by the goto.com search suggestion tool for the same keyword last month. This is only on a single keyword though, in a specific industry.
I think the best thing to do is to use the search suggestion tool to see how many people searched for your keyword phrase last month, and then test an Adwords ad on the same keyword phrase to research a rough coorelation and prediction of how many adwords impressions to expect in your industry.
Let me know if this answers your question.
That's what prompted me to ask the question. I put in "internet security" into the keyword tracker>> On GoTo it shows 17,452 searches done in August, (which is admittedly fairly low compared to huge keywords like "music lyrics" which had 169,771.) the top bidder pays 59 cents for it.
On Google Adwords it came up as 105,400 impressions. And on WordTracker it came up as 1876.
I just don't know how to compare these numbers because they are clearly not "apples vs apples". They use different means of getting the data, different months even maybe.
I just don't want to make a play for a keyword or phrase that hardly anyone ever types in.
I really appreciate your help in this matter by the way.
Well what really matters is clickthrough rates, IMHO. If you target a more generalized phrase with tons of searches, but nobody clicks through, you are just burning dollars.
You can put a single ad up on any number of keyword searches, so I suggest trying all the decent ones that the keyword suggestion tool provides. You will find out which are getting good clickthrough rates (low cost per clicks), and which you are wasting your money on. If you keep adding terms and revising to keep only high click through stuff, you will eventually meet your spending limit and you will be getting the most bang for your buck!
Unless your goal is purely branding, that is. :)
We talked a bit about comparing numbers in the difficulty in coming up with an accurate global prediction in this thread.
The main problem you're running into is that you're not comparing similar things.
1. WT Count vs Predict
The number you gave from WT was the Count (number of occurences in the sample database)The Predict (estimation of daily searches across all major engines) is 1911.
2. AdWord Limitations
The AdWords tool states that "Estimates of how many ads will be shown and its cost are based on past history, do not take language or country targeting preferences into account, do not distinguish between 'Keyword/ Phrase' and 'Exact Query' preferences"
That means that even though you might have put "internet security" into the Exact Phrase box, it is still giving you a count for all searches that contained the two words. (If you paste your phrase into each of the boxes and run them, you'll see how the estimates it returns are all the same).
The Count number from WT was only for the specific phrase "internet security." In order to conduct any kind of comparison, you would need to collect every single search in WT's databse that contained your phrase and then up.
3. Factor In Automated Reporting
Both Google's and Goto's counts will include queries conducted by automated reporting software, while WT's will not. I don't know of any way to calculate the percentage, but you will always find large differences between WT and Google/GoTo count numbers when comparing very popular terms. When you compare more obscure terms, (ones that are much less likely to be included in position reports) the numbers are much closer.
4. Plurals & Mispellings
GoTo also lumps all singular,plural and misspelled variations together, which makes it a bit difficult to use it to make any type of global prediction.
I read your post from Sept 01
> The Count number from WT was only for the specific phrase "internet security."
I have just been looking for that within wordtracker. The touble with that is then you can only compare exact term searches with exact term competing web pages and this in my opinion is not how users use the internet search engines. It certainly is not how I use them, I just type in what I want and see what I get I rarely use advanced options and very rarely use quotes.
So it seems to me at the moment that the only valid wordtracker competition reports are for exact phrase against web page competing for exact phrase.
For many niche key phrases this shows no competing pages yet when excluding the "" and cheking the search engine directly you find there are tens of thousands of competing pages and sites.
As I think users search mainly without the "" it devalues the wordtracker results and their KEI estimation especially.
What do you think?
The reason WT uses quotes, and you should also use quotes when looking for competition is this: You are looking for competition on that specific phrase. If "keyword phrase" returns 0 competing pages, that means no one is targeting that specific phrase, and you should be able to obtain rankings for it fairly easily. Searching for keyword phrase without quotes will likely produce more listings, as they can have keyword and phrase on the page, but not necessarily together. They aren't (necessarliy) optimizing for the exact phrase, which is what you should be trying to guage.
Waiting for a clear response from customer service on a question.
"If "keyword phrase" returns 0 competing pages, that means no one is targeting that specific phrase..."
Drastic, wouldn't that also suggest that few if any searchers are using that phrase? I can see the point of finding phrases that other sites aren't targeting, but only if I can be sure people are using those phrases in searches.
I have several #1 and #2 positions for phrases that I thought people would use. However, they aren't bringing in much traffic. I can see in my logs what terms were used when people found me, but what I need to know is what terms were used when they didn't find me.
>wouldn't that also suggest that few if any searchers are using that phrase?
Not necessarily, it just means it hasn't been exploited yet. You still have to do your research for popularity of a phrase. 0 competing pages is not really a realistic figure for a good phrase, I was using it as an example. (you won't find many gems like this - if/when you do, jump on it.) You want to find a good ratio of competition to popularity. What I usually do is get a list of popular keywords, and then check the competition on them.
Regarding WT, I'm assuming that by the time you get to the point where you need to decide whether or not to run a competitive check using quotes, you have already collected a list of terms that shows a decent level of search activity.
Expanding a bit on Drastic's comments, using quotes will also help you determine how many individual pages of content you will need to gain visibility for a set of related terms.
If I have a group of 10 related phrases comes back with three phrases that have a high number of competing pages for an exact match, and seven that don't, then I know that I'll probably only need to develop/optimize three pages. The highly competitive phrases each get their own page with their own title and descriptions. The other seven terms can usually be worked in by creating close proximity matches within the copy.