13 replies to this topic

[JamesW]

Hey,

Does anyone use a mathematic formula or equation to decide on their potential keyphrases?

For example, assume the following searches led to these results:

(1)red widgets (10,000 pages)
(2)dark red widgets (5,000 pages)
(3)reddish widgets (1,000 pages)
(4)reddy colour widget (500 pages)
(5)widgets with a little bit of red in them (250 pages)

Now working on the basis that you choose three keyphrases to target per page, and assume these to be very accurate keyphrases for the page in question, you input the phrases into your Wordtracker account/other search estimate software/etc.

The following results appear for the number of searches conducted per week:

(1)red widgets (5,000)
(2)dark red widgets (2,500)
(3)reddish widgets (500)
(4)reddy colour widgets (250)
(5)Widgets with a little bit of red in them (175)

On the basis of these (rather improbable) results from your research (which we assume to be consistent on a weekly basis), which three keyphrases would you choose? Does it matter?

Would your decision change if the following results for the number of searches appeared:

(1)red widgets (4,000) (10,000/4,000=2.5)
(2)dark red widgets (2,800) (5,000/2,800=1.78)
(3)reddish widgets (500) (1,0000/500=2)
(4)reddy colour widgets (300) (500/300=1.66)
(5)Widgets with a little bit of red in them (200) (300/200=1.5)

Would you choose No.5, No.4 and No.2? Is the traffic really worth using keyword #5? What do you think? Is it right suggesting that there are 3 types of SEO techniques:

1) Blind ambition - go for the 'top' (whatever they decide as top) three keywords
2) Fingers in many pies - Going for the top, middle and bottom keyphrases
3) Conservative - Choosing the last three keyphrases

Which one of these, if any, do you fit into?

P.S. If you've selected 'red widgets' and 'widgets red' as your keyphrases, is there any need to include both., as there appears to be the same number of pages on index, with similar top 10 rankings for both.

Cheers,

James

[compar]

Would your decision change if the following results for the number of searches appeared:

(1)red widgets (4,000) (10,000/4,000=2.5)
(2)dark red widgets (2,800) (5,000/2,800=1.78)
(3)reddish widgets (500) (1,0000/500=2)
(4)reddy colour widgets (300) (500/300=1.66)
(5)Widgets with a little bit of red in them (200) (300/200=1.5)

Would you choose No.5, No.4  and No.2? Is the traffic really worth using keyword #5? What do you think? Is it right suggesting that there are 3 types of SEO techniques:

I'm not sure what you think the calculated number indicates. As I understand it it is the total pages in the SERP for the particular keyword divided by the the number of searches in some time period.

What do you think this gives you? A competition ranking like WordTracker's KEI? You seem to suggest that possibly the keywords with the lower calculated number are preferable. Can you explain the logic behind this?

[JamesW]

Bob,

The maths are simply a way of showing the searches-to-pages ratio. The lower numbers, in theory, provide the better searches as these are the best search-to-page ratio. This is not my particular view, but is a discussion of a possible mathematical (if not scientific) reason for choosing keywords.
Cheers,

James

[susan]

Personally, I think I would pick the #1, #2 and #5 choices. I don't want to be left out of the chance of coming up under #1, and at the same time would like to take advantage of coming up (perhaps) in the top of the #5s.

[JamesW]

Susan,

Is it really worth it for the limited amount of traffic that the last phrase would give you? After all, should be relatively easy to gain good rankings for this phrase - are you not perhaps wasting your talents?

Cheers,

James (aka Devil's Advocat)

[compar]

Is it really worth it for the limited amount of traffic that the last phrase would give you? After all, should be relatively easy to gain good rankings for this phrase - are you not perhaps wasting your talents?

As has been said elsewhere in this forum that may depend on the value of the item being sold and the profit per sale. If you only need to sell 2 or 3 items a week #5 may be just fine. If you need to sell a bunch of items every day then you had better go for #1.

So I don't think you can make the "are you wasting your talent" descision without more fact about the page and the nature of the product.

[Scottie]

Welcome to the forum Susan!

I believe that keyword research is part science, part gut feeling. The better you or your client know the market, the stronger your "gut" will be for certain phrases. It never hurts to poll some potential customers to find out how they would go about searching for your product or service. (always testing... my poor friends and family ) Often, the results are surprising.

There is an art to combining certain phrases on a page in order to maximize the potential phrases that page will rank well for.

It's also interesting to play with Alta Vista's Prism tool that suggests related terms.

[BrianR]

Scottie

Couldn't agree more about research. Especially in technical areas, asking staff and customers yields niche keyphrases that you'd otherwise miss.

Marketing axiom: When all else fails, ask the customer!

BrianR

[compar]

Scottie

Couldn't agree more about research. Especially in technical areas, asking staff and customers yields niche keyphrases that you'd otherwise miss.

HBrian,

What you and Scottie say is very true, but has almost nothing to do with the question of this thread. The question was all about using arithmetic to select which keyword phrases to use. Talking to you customers has nothing to do with the question.

Good advice though.

[Scottie]

Well Bob, the title is:

Keywords Discussed, Is there a science to deciding?

So I listed my opinion that deciding on keywords is part science, part gut feeling, part user-research. If that's not on-topic, I must have missed something.

Aside from that, if we were to hold every thread to a strict, narrow interpretation of the original question, we'd miss a lot of interesting discussion, don't you think?

[Haystack]

Hi James.

In situations like this, I target #2 on the front page of the site, which happens to cover #1. Then I'd use sub pages to go after the less competitive terms.

[DanThies]

I'd use Adwords and Clicktracks to figure out which of the candidates is best targeted, and most likely to result in revenue/profits. *Then* I'd do the rest of the math, to figure out what the "payoff" would be for successful optimization efforts, and go after the biggest bags of cash first.

[Mel]

Why do you think that the number of pages that have been published on a particular topic has any bearing on what keywords you should choose?

IMO the main reason behind keyword research is to discover how searchers are searching for keywords related to your site and how many pages published is a only a matter of how popular (or maybe profitable) the topic is.

My answer to the question is to totally forget about how many pages are published if you want to decide on keywords. Look (really look) at how your potential viewers are using keyphrases for search relative to your topic and then look at the competition, but only those on the first page of the SERP.

[awall19]

Other pages with the term is essentially an irrelevant data point when it comes to SEO. The # of well optimized pages is what really matters. I believe the most reasonable targeted term is the home page and the inner pages should try for other terms.

I really agree with haystacks idea of doing #2 if it includes #1. This benifits in three ways.

1.) you get # 2
2.) you potentially get #1
3.) #2 can be used as a barameter to help capture phrase set #1. kind of think of it as futures or as a leading economic (targeted traffic) indicator.