63 replies to this topic

[David1221]

Hello Jill and Dan,

I have a very interesting scenario which you guys may be able to explain. I am a current user of Wordtracker and Keyword Discovery. At the moment I am using KD as a testing tool before I actually implement it into my actual campaigns. The problem is that they both represent extremely different results for some cases.

An example is as follows:

I typed a few keywords into WT and KD, and I would get the following results -

WT returned 500 searches per day
KD returned 200 searches per 12 months

From this information, I would either think that the following keywords are obviously very low search term keywords (based on KD) or high search term keywords (based on WT). Then my gut instinct would tell me that it is a keyword entry that people would enter into the search engine... but how can these results be so far apart? What advice would you give, is it a low search term or a high search term? And lastly, which is more accurate?

At this stage I would tend to believe Wordtracker over KD since KD has just hit the scene... and when I confronted a help officer from KD, they suggested that the keywords in KD are actual keywords in their database (over 12 months worth of searches) that have been searched from 180 search engines.... the irony..... They have more search engines to obtain data from than Wordtracker yet they only come up with 200 searches in 12 months!!!

Any input is much appreciated.

Regards,

David

[zacheos]

I like word tracker... though neither are perfect and are only return an estimate - The more widely searched a phrase is, the more accurate the reports will be. No reason not to use several methods of getting an estimate though. If you can afford it, I recommend running several trial accounts with a variety of PPC engines to test out certain phrases and get a sense of the market.

[Scottie]

They use different sources for their data.

You can't use any keyword research tool as an absolute- none of them measure all actual searches done in the world.

What they are excellent for is for benchmarking, relative popularity, and alternate phrases.

I'm addicted to Keyword Discovery- I find myself logging in for a quick check to something and hours later, I'm still gathering data...

[davidfromraleigh]

I'm with Scottie. I love Keyword Discovery. It's so easy to just log in and get started. As for the results, Scottie covered that as well because they are just a benchmark yet very helpful.

David

[David1221]

Here is an update on my research.

I think Keyword Discovery is an absolute waste of money and I will be asking for a complete refund.

Why?

Well here are the results and then you be the judge:

For a word like "online website marketing", WT predict a 112 search/day..... To me that sounds very reasonable since this keyword is sensible.

The results in KD show 233 searches per year! Now you don't need to be a scientist to work out the logic of this. I would say 233 searches / month is low but could be possible, however 233 searches per YEAR, now that can't be right especially considering that they take results from over 180 search engines.

You would simply rack up 233 searches in a month quiet easily, considering that SEO optimisers would be looking to possibly optimised for that term therefore entering it into search engines to check competition. Even Overture returns more results than KD for that keyword, and again, that is on a monthly basis.

I can understand if this happened in a few instances, but I have typed in other keywords and the results are similar to the example above.

I think it is back to the drawing boards for KD as they have a lot of catching up to do.... especially considering that they are charging you quiet a bit of money for this service.

Any thoughts?


David

[projectphp]

120 searches a day for [online website marketing]? I doubt it.

look at http://inventory.ove...bsite marketing. That shows a few a month.

And the question is: how many are real searches, and how many automated?

Personally, in judging a tool, I would use something less driven by small sample size. Something like "flowers" or "tiger woods" would bea fairer indication. These are terms that get massive search, and are less affected by rank checkers than are terms that online marketers target.

[David1221]

Ok.... lets go with Overture...... minus all the automated searches.... lets be strict and say 200 searches a month with Overture...... How in the world did KD come up with 233 searches PER YEAR!!!

If we say with overture 200 per month and they only look at a few search engines, then the yearly figure would be about 2,400 searches..... On the other hand, KD uses 180 search engines including the ones Overture uses and they can only come up with 233 searches.......... that ain't right!

[seadog]

David,
I think you'll find it a better use of your time to see how many people "convert" on the site you are optimizing for the keyword rather than how many people searched for it. Even if 1,000,000 people searched for a term its hard to tell if they will convert until you implement something. As mentioned before PPC is usually a good way to test these words

[projectphp]

I have no idea how they came up with that estimate.

Now, how you spend you $$ is a question only you can answer, but I wonder if the test you have is an accurate one. I would say al of the tools are not that accurate, and it comes down to what you personally decide is best.

[RiYo]

Is there not somebody that uses KD as well to check if the numbers that David comes up with are correct? Perhaps he did something wrong. The HUGE difference between the 2 does not make sense, I know that you better stick to one method, but difference is a bit too much.

Richard

[David1221]

Thanks guys for all your input.... the point i am trying to make here is that KD are charging an amount of money and the results it gives are way off.......

I even confronted the help desk on a live chat and he/her did the research for themself.... and let me tell you it, it took him/her a good 20 minutes to get back to me cause he/her needed a justified explanation for herslef/himself from the manager....

At the end of the day, it is important to make sure you get the conversions but to be able to get some free traffic in the first place, you need to have good keywords...

With all that said.... PPC would be the best way to test if you have the $$ but for now i think wordtracker is the tool i will stick with along with my gut instinct.

I would appreciate anyone else who can maybe find some reason for this strange outcome of results :-)

[RiYo]

David, what was the reaction of the manager?

[Jill]

Actually, I would believe it. Who (real people) would search for online website marketing?

I can't imagine EVER a real person searching for that. So 233 a year...yep, a few of those people checking their rankings manually could slip through.

With keyword research, use your common sense. Why would a person ever use the word "online" with "website marketing"? They wouldn't, because it's obviously online. It's only those silly web marketers themselves who needed to add a phrase in that they might actually be able to rank highly for who use it.

Betcha you'd never get any traffic with that phrase if you were number 1 in google. Not very much at all.

[AlDugan]

You also have to keep in mind that the Overture Keyword Suggestion tool will always be higher than a tool such as Wordtracker or KD... for "online website marketing" as the example, Overture will also count the following terms as that same term:

• website marketing online
• web site marketing online
• online website marketing
• online web site marketing
• website online marketing
• web site online marketing

You get the idea...

[rcavilla]

If you click analyze for "online web site marketing" on KD you get an estimated daily of 5 per day, multiply that out you get 150 per month, which for this keyword phrase I think would be in line- so don't just look at the "searches" it may be misleading.

You've got to remember the sample sizes that each tool is working with. Try to back the number out using the tool. For instance KD tracks 26 billion searches over a year, use your favorite industry statistic for estimated total searches per month / year, get the ratio to 12 billion then multiply the ratio by the initial "searches" figure.

Regarding KD- be careful as I think they may count a root word every time it appears in a phrase. For example:
KD for "hair loss": 6,573,493 searches or 150,080 per day
WTracker for "hair loss": 1,859 daily

So in many cases KD is much, much, higher- my thought is that its counting "hair loss" every time it sees "hair loss solution", "hair loss treatment" etc.

I find sometimes wordtracker will leave out obvious keywords, so i start with wordtracker then go to keyword discovery to continue building out the list. Just my two cents.