homepage Welcome to WebmasterWorld Guest from 23.22.128.96
register, free tools, login, search, pro membership, help, library, announcements, recent posts, open posts,
Become a Pro Member
Visit PubCon.com
Home / Forums Index / Marketing and Biz Dev / Keyword Discussion
Forum Library, Charter, Moderator: open

Keyword Discussion Forum

    
finding "good" keywords
is there a good method for finding and then fine-tuning keywords?
leanweb




msg:269502
 4:40 pm on Jul 6, 2003 (gmt 0)

continuing saga of red widgets.

so i sell red widgets. but there are 10,000 other websites that sell red widgets. so if my primary keywords were "red widgets", then it could be tough getting on the first SERP.

but i dont know what people look for when they look for red widgets. perhaps yellow widgets? or red gadgets? i'll have no way of knowing, will i?

so i guess my question is, would ya all good folks share tips/ideas on how to mine for good keywords? i mean practical stuff, even to trial and error madness there has to be a method, no? ;)

thanks in advance,
leanweb

 

SEO practioner




msg:269503
 4:54 pm on Jul 6, 2003 (gmt 0)

Hello leanweb and welcome to WebmasterWorld!

First, are you using wordtracker?

Second, try 3 word keywords, such as:

* Large red widgets

* Small red widgets

* Medium red widgets

... etc.

3 word keywords are easier to optimize as they have less competition. By the same token, 4 word keywords are even easier, as they get even less competition.

Good luck

Shak




msg:269504
 4:58 pm on Jul 6, 2003 (gmt 0)

PPC inventory tools are always a good idea, I am assuming you know about these?

if not here is the biggest of them all:

[inventory.overture.com...]

Shak

Lorel




msg:269505
 5:00 pm on Jul 6, 2003 (gmt 0)

Hi Lean,

>but i dont know what people look for when they look for red widgets. perhaps yellow widgets? or red gadgets? i'll have no way of knowing, will i?

>so i guess my question is, would ya all good folks share tips/ideas on how to mine for good keywords? i mean practical stuff, even to trial and error madness there has to be a method, no? ;)

I have a method that is working well for the sites I have designed. First I would put red, yellow and whatever color widgets you think will sell in the text of the page. Then, after Google or other engines have had a chance to refresh your page, I would watch the counter referrals (search engines linking to your site listing the search term) and whenever one of those words come up I click on it to see how it ranked (and just the fact that someone searched for that word means it has to be fairly high in the ranking on that engine). Then if that is one of my major keywords then I make sure that page is optimized for that word, and maybe add it once more if it didn't get top billing.

I also check Overture's Search Term Suggestion tool to see if there is some other way I should be using that word(s)--a phrase will gain you more hits than a single word. Sometimes a phrase gets more hits when it's turned around, i.e., "widgets green" instead of "green widgets".

I constantly fine tune my web sites by what I find in my counter/site meter and my sites are all gaining in popularity in major engines. And in fact I have several that are reaching top billing in major engines and I've never paid for a listing in any of them--even Overture.

Shak




msg:269506
 5:04 pm on Jul 6, 2003 (gmt 0)

I also check Overture's Search Term Suggestion tool to see if there is some other way I should be using that word(s)--a phrase will gain you more hits than a single word. Sometimes a phrase gets more hits when it's turned around, i.e., "widgets green" instead of "green widgets".

I would disagree with the above, as that is Overture matchdriver in place.

doing a search on OV tool brings up some weird numbers like you mentioned, but those are NOT 100% correct...

Shak

leanweb




msg:269507
 5:28 pm on Jul 6, 2003 (gmt 0)

(sorry for posting on the wrong board)

good stuff!

next step is, how can i identify the best converting keywords?

ideally, of course, it would be great to know keyword conversion potential in advance, before i include them on my site, although this seems like to tough of a task.

is it possible to track which keywords result in a sale? are there tracking tools for that?

looking forward to more practical advice!
leanweb

SEO practioner




msg:269508
 5:42 pm on Jul 6, 2003 (gmt 0)

"is it possible to track which keywords result in a sale? are there tracking tools for that?"

go to [wordtracker.com...]

its the best

Lorel




msg:269509
 6:12 pm on Jul 6, 2003 (gmt 0)

Hi Shak,

I had said:
>>I also check Overture's Search Term Suggestion tool to see if there is some other way I should be using that word(s)--a phrase will gain you more hits than a single word. Sometimes a phrase gets more hits when it's turned around, i.e., "widgets green" instead of "green widgets".

>I would disagree with the above, as that is Overture matchdriver in place.

Which part of what I said do you disagree with--using Overture's search term suggestion tool or using words backwards? I think both are a good idea (weighed with common sense of course) and particularly the later because common words like "in", i.e., "widgest IN green" will not be used by the search engines.

So please explain.

BTW, I find Overture's Search Term Suggestion Tool much easier to use and more flexible than Word Tracker and I would never pay for the use of the later when the former is free.

sem4u




msg:269510
 6:18 pm on Jul 6, 2003 (gmt 0)

Overture's Search Term Suggestion tool counts "widgets green" as the same keyphrase as "green widget". Therefore only one keyphrase will appear in the suggestions list, inflating the figures.

Shak




msg:269511
 6:38 pm on Jul 6, 2003 (gmt 0)

I think sem4u just answered the question in relation to matchdriver and OV.

"is it possible to track which keywords result in a sale? are there tracking tools for that?"
go to [wordtracker.com...]

Last I looked, Wordtracker was NOT doing ROI tracking on specific keywords resulting in Conversion rates.

The ONLY way of working that out is Trial n Error, combined with the usability of your site and other factors such as price, demand etc etc.

Wordtracker is good for sourcing inventory to a point, but NOT for defining which keywrods convert better...

Shak

leanweb




msg:269512
 8:45 pm on Jul 6, 2003 (gmt 0)

so, in other words, there isnt a tool to track conversion potential for specific keywords?

any tips on how to go about estimating?

let me give you an example of why i think this is important.

in the last few days traffic on my site spiked up at least three fold. reason? 4th of july. people were looking for fireworks. i have the word "fireworks" on my site here and there. naturally none of the folks looking for this explosive product were interested in my "red widgets." that was completely useless traffic.

no big deal you say? i think it is. waste of bandwith is least of my worries, of course. what worries me is feeling of comlacency that can be enstilled by high (yet useless) traffic figures. also, i dont want to waste time optimizing for keywords that bring no sales.

bottom line, whats the point of optimizing for keyword if it has no or little conversion rate?

question still stands: how to go about estimating conversion potential for individual keywords? if not tools, then methods may be. this may be obvious, but indulge me, a beginner, throw few pointers my way. please? ;)

thank you in advance,
leanweb

buckworks




msg:269513
 10:13 pm on Jul 6, 2003 (gmt 0)

<<how to go about estimating conversion potential for individual keywords?>>

There are no shortcuts here.

And if there were, people would be unlikely to give away secrets that they considered to be a competitive advantage.

Do your best to get into the mind of the searcher ... what would a person be thinking when they searched for this phrase? Is that who I want to reach? Make your best guess about what will be effective in your context, then test it. If it works, keep doing it; if it doesn't, try something different. Progress is a never-ending process of testing and tweaking, discarding what doesn't work and developing what does.

leanweb




msg:269514
 1:02 am on Jul 7, 2003 (gmt 0)

buckworks,

thanks for confirming my hypothesis.

is it fair to conclude then, that the crooks of the matter as far as SEO goes is finding which keywords sell best? the rest seems less of a challenge, given the amount of info on the subject.

finding good converting keywords is highly guarded trade secret. so be it.

having said that, are there any baseline metrics for what conversion rates are "healthy"?

i do realize that conversion rate would depend on what these widgets are and their price. but may be you can offer an example?

thanks again in advance,
leanweb

fom2001uk




msg:269515
 4:35 pm on Jul 7, 2003 (gmt 0)

Leanweb, there ARE ways of tracking keyword conversions. Go visit the tracking and logging forum :-)

Robert Charlton




msg:269516
 10:39 pm on Jul 7, 2003 (gmt 0)

how can i identify the best converting keywords?

GoToast is one such tool... web-based, using javascript tracking code on each page. It can track PPC campaigns and search engine referrals with search string information. It uses cookies to identify return visitors, etc. They charge a monthly fee, plus overages if you exceed your allotted monthly clicks. There are other such services too.

You can also set up an AdWords campaign, using exact phrase matching, to see which phrases convert.

There are liable to be all sorts of factors in addition to the phrases themselves that you'd need to consider in evaluating any conversion data... including ad copy, ad position, etc. In PPC conversion evaluation, remember you're only evaluating within the universe of what you do target.

I've seen clients jump to wrong conclusions about what phrases convert based on small sample sizes as well as inadequate target universes. I think you have to observe behavior over a period of time with a bunch of data before conversion data makes sense. Others more experienced than I in tracking conversions may have other opinions. I would check the tracking forum.

In general, the more specific a phrase is, the more you can perhaps assume the searcher knows what he's looking for. From this, I generally assume that long phrases are more likely to convert, but they're also less likely to be searched.

Overture data (and WordTracker data too), can be skewed by automated searches, so I'd interpret all such data skeptically.

Alpicat




msg:269517
 6:50 pm on Jul 10, 2003 (gmt 0)

Hi

Have you thought about setting up a survey on your site? I've tried it and got very interesting results - suprising the number of people that will help out if there's an incentive and combined with a few other key questions you'll get an insight into how people find your site and the areas they see for improvement.

Cheers
Andy

JoeHouse




msg:269518
 11:11 pm on Jul 13, 2003 (gmt 0)

leanweb

Some of my best and sucessful keywords are keywords I haven't even selected. If you have a well designed website full of great content and relevancy to what you sell the results will surprise you.

For Example:

If you sell widgets and somewhere on your homepage you mention orange widgets with bows and ribbons etc, etc, etc...

When a search is done on widgets but not your chosen keywords but they are in your description somewhere(preferably on your homepage), your site still has a very good chance of being found on that particular search result and ranked highly too. But it really depends on how well google likes your site and how spiderable friendly it is.

Hope that helps.

emailtools




msg:269519
 11:43 pm on Jul 13, 2003 (gmt 0)

leanweb

You could try a quick AdWords campaign (say $50 worth), using broad keyword matches only. Just track the keyphrases that send you traffic and log any sales by keyword too. You'll have a pretty accurate idea of what keywords will bring targeted traffic to your site in no time at all!

You'll need to use a tracking tool to log the data. Sticky me if you'd like to know the one I recommend to clients.

Toasted




msg:269520
 10:25 pm on Jul 19, 2003 (gmt 0)

As a general rule, the more targeted the keyword is towards your traffic, the higher the conversion rate will be.

For example - if you sell "Bright Red Widgets with Orange Spots', you might get a lot of traffic for 'Red Widgets', but it might not convert too well (depending on how common that term is compared to how speciaized the variant is you are selling.

'Bright Red Widgets' should get you a higer conversion rate, 'Bright Red Widgets Orange Spots' higher still.

The other thing you have to watch here, is the more you zero in on the full term, the less traffic you'll get - so even though the conversion rate will be higher, the total number of sales (and/or ROI) may be lower.

Other than that, the only way you can see which keywords convert, is to track them yourself (as others have pointed out)... there's nothing out there (that I'm aware of) that will tell you some magical formula as to what keywords have good conversion rates, because what converts for one person, might not convert for others - to use your example earlier, if you sell 'Red Widgets' then searches for 'Fireworks' probably won't convert... but if you were selling Fireworks, then I think it might convert reasonably well (though it could probably still do with a little more targeting).

There is one other way you can get a general idea I guess - have a look to see how much people are bidding for the term on Overture - the higher the bids, the (probably) better conversion... this could be a little misleading though, as competition for a term does not necessarily mean better conversion - but you know someone is making money from the term at least.

CromeYellow




msg:269521
 10:03 pm on Jul 29, 2003 (gmt 0)

Hi leanweb

We just had a programmer knock up a keyword conversion script and can now see keyword, search engine, entry page etc etc for every sale. It cost a few hundred bucks and is worth its weight in gold. :) I'm sure at Elance you could get someone to do you one even cheaper.

Cy

aspdaddy




msg:269522
 10:54 pm on Jul 29, 2003 (gmt 0)

I would also suggest mining any existing log files, customer emails and customer websites.

An internal site search is another way to get insight into what keywords your customers use.

You may find new angles that you wouldnt always get from using the main search suggestion tools.

Global Options:
 top home search open messages active posts  
 

Home / Forums Index / Marketing and Biz Dev / Keyword Discussion
rss feed

All trademarks and copyrights held by respective owners. Member comments are owned by the poster.
Home ¦ Free Tools ¦ Terms of Service ¦ Privacy Policy ¦ Report Problem ¦ About ¦ Library ¦ Newsletter
WebmasterWorld is a Developer Shed Community owned by Jim Boykin.
© Webmaster World 1996-2014 all rights reserved