They've done their campaign entirely with broad matching, and some of the matches overlap. Eg, they bought, both "red widget" and "red widget sales."
Can I assume here that there are no "red widget sales" matches in the broad matched "red widget" click-thrus, or is this not clear cut?
I'm guessing I will have to go back and have them do a campaign with exact targeting for "red widgets," "red widget," "red widget sales," etc, to get more useful data, but I'd like to try to get something out of their original campaign... and, besides that, I'm curious about it.
Can I assume here that there are no "red widget sales" matches in the broad matched "red widget" click-thrus,
That is correct, assuming that the Max CPC of the broad term doesn't exceed that of the more specific term.
That is correct, assuming that the Max CPC of the broad term doesn't exceed that of the more specific term.
Hmmmm.... I've been mulling this one over. I'm not familiar with the subtleties of AdWords, though, so I'm not quite making sense of your answer.
The spreadsheet I have shows that the Max CPC of both phrases is the same, and that the average CPC of the more specific term is actually double the average CPC of the broader term (what I'm comparing here are figures for the two-word phrase and the three-word phrase, both broadly targeted).
This goes against my intuition, as in regular SEO it's harder to rank on a broader term, so the broader term feels more valuable, but I think I could make an argument for a pricing structure that works this way.
What is the factor that takes precedence... specificity or cost? If cost, since I'm looking at average figures only, how do I tell? Beyond pinning down the answer to my question, I'd appreciate whatever philosophical comments on the pricing structure you'd care to make... ;)
that the average CPC of the more specific term is actually double the average CPC of the broader term
Well, that depends on who is bidding and what they set their max cpc to.
You're correct in assuming that the broader terms could be more competitive. You have a larger pool of bidders with possible overlap between many industries. However, the max cpc's for the different sets of bidders varies.
So, if someone searches on "Blue Widget Software" you get one set of bidders. If someone searches for "Dentist Widget Software" you get a different set of bidders. The overlap creates a dynamic situation where the pool of bidders grows and shrinks, thus affecting the "average" bid.
Let's assume you have a max bid of $.55.
For a search on Dentist widget software: If there are FOUR BIDDERS, and you are appearing at position three, the bidderr in position TWO has a max cpc higher than $.55.
But, the max cpc of the bidder beneath you, at position four, is $.50. Then this creates the situation of you having to bid .51 to maintain above the bidder beneath you.
But suppose the searcher types in BOOK Widget Software? A different set of bidders comes into play, who are less competitive. Let's assume that the bidder beneath you has set their max cpc at .10 and the bidder above you has set their max cpc at .56, then your cpc for the term BOOK Widget Software will be lower, eleven cents.
BUT, if your term is more specific, let's say you are geo-targeting for your city. Furthermore, let's assume that you have many competitors here who assume that the more specific term provides a better conversion rate.
The better conversion rate makes the more specific term more valuable.
Ironically, ranking on some of the more specific three-word phrases is a no brainer, whereas several of the two-word phrases are a little bit competitive.
