54 replies to this topic

[janell]

Hi all,

In doing research in KD, I've noticed quite a few keyphrases that just seem to be in a mixed up order, kind of like how I might expect to see them in Overture. But I'm specifically not searching in Overture.

I was under the impression that the KD database left the words in the correct order, based on what the user originally searched for. Is that not the case? Are some of the phrases that we get out of the regular KD database of words also going to be out of order?

(To share the specific example, I'm seeing 409 searches for "based business cleaning home office" and it certainly doesn't make sense for based to be the first word in that phrase for that many searches.)

I guess I'd just like to know if I should expect that KD results are going to give me phrases with the words out of order?

Thanks for your help,
Janell

[Jill]

Funny you should say that.

I was actually doing a live KD demo today at our seminar in front of a whole classroom of people, and I was seeing really weird results as well.

Obvious automated query stuff was showing up really highly. I used it as a good way to talk about using common sense when choosing your keywords.

But I did wonder why all of a sudden automated query weird words were creeping in as I haven't seen as much of that previously.

It wasn't really a good time to see it as I was trying to show how great KD was in the seminar though!

[DanThies]

It looks like Overture data to me, since they like to put the words in alphabetical order. Guess we gotta wait for Dave or Ren or someone to stop by and let us know.

[Trellian]

Hi All,

You know we have always had those strange terms in our database. The problem is the more search engine data we collect form the main search engines (Google and Yahoo) which is more and more now... the more we find our data skewed.

So basically we are at a juncture...
do we not include any top tier search engine data and keep the data cleaner,
or
do we include the top tier engine keyword data to provide a more accurate perspective of what you are likely to receive in terms of searches on Google and Yahoo if you are running PPC campaigns on these 2 engines.

The answer to that still remains unanswered.

One option we could look at doing is maybe we look at providing another database where we just include keyword terms from ISP files, user panels, meta engines which are less prone to skews as a separate database... Closer to a wordtracker service that just includes keyword data from meta engines.

Maybe send me your thoughts as to what you all prefer..

[Jill]

I dunno...but this stuff definitely didn't used to be there, so I would vote for going back to however it was before!

[Trellian]

It was always there Jill,

There is just a bit more now with the growth of our data imports.

Cheers
David

[Randy]

hmm... If I had a vote I would cast it in favor of giving the user some way to separate out the data that is causing the skew. That way people could potentially have the best of both worlds.

The only problem would be in how to explain the difference so that the average person could understand what's what and why.

[Jill]

Well, Randy, you just don't have to click on and save the results that don't look right, but it's a judgement call.

I just don't recall seeing this much of it so high up in the KD results before. I was somewhat embarrassed while doing my presentation. since I am always pushing it's not generally effected by automated queries.

[janell]

Hmmm.... I agree with Randy, it would be nice to be able to have both options. Just because I like options. You can always find a use for them.

But seriously, I can see wanting to leave out those kinds of results while gathering a bunch of possible keywords, but wanting to see them when you're comparing numbers on specific phrases.

My question is this - are they valid searches with a mixed up word order, or automated searches that should be completely ignored?

Janell

[Trellian]

My question is this - are they valid searches with a mixed up word order, or automated searches that should be completely ignored?

If you are doing Organic SEO, then they are not valid.
But if you are running Google AdWords or running Yahoo/Overture PPC campaigns, then they are very much valid and you will get traffic for these terms.

Hence the dilemma, for some users this is useful to know, for others it is not.

Cheers
David

[Jill]

Well, as an organic SEO (no pesticides here) then they're not useful. Looks like you need a button for organic and one for those who use pesticides perhaps, David?

[janell]

lol, that sounds like a good way to explain them, organic vs.

[DanThies]

Definitely would help to have some way to separate the two, David. It used to be more rare to see these kinds of terms, but it's never easy to separate them out by hand.

[Trellian]

Dan and Jill,

Check now. Yesterday we did a major update on the KeywordDiscovery keyword database... which also grew from 26 billion to 32 billion searches, since our last large update.

Cheers
David

[Jill]

Will let you know, David! Thanks!