An exact match with a higher bid will get more clicks.
A phrase match will show after the exact matches, especially if your ctr is lower.
Generally speaking our best conversions are on exact matches, we have many examples of way cheaper phrase match traffic, but we cannot say for sure if the keywords would show to the same extent as with the exact match, if that makes any sense.
We have adgroups that have exact, phrase and broad, but I imagine it would be better splitting these into 3 ad groups on big volume kws because broad match can often bring the ctr down which, if they shared the same adtext, would affect the exact match.
Hope that makes sense.
Obviously, [green widgets] can only show for [green widgets] EXACTLY and nothing else. "green widgets" can show for any query that has "green widgets" in it.
So there's only 1 area of possible overlap - when the searcher types in [green widgets] precisely... The question is.. will that impression get allocated to [green widgets] or "green widgets"?
As I have had it explained to me, each term would be evaluated for its quality ranking and bid combination, which would be affected by the click-through history of each. During the overlap scenarios, the higher scoring term would be shown. If all things are exactly equal (like both terms being brand new), the traffic would be split 50/50.
Theoretically, [green widgets] would show... because, historically, less targeted phrase match traffic would have been weeded out - producing typically a significantly higher CTR and QS than "green widgets"...
[webmasterworld.com...]
Here's a thread where eWhisper tells AWA that it's actually not working like that in his account and AWA insisting that things work like he said :)
[webmasterworld.com...]
Waa.. what a mess.
He said: if there's [apples] and "apples" in the same adgroup and searcher searches for precisely 'apples', then ALWAYS [apples] will get the impression. That's what he said.
Now what you just said is COMPLETELY different!
"If multiple keywords from your account could potentially be triggered by the search query, the keyword which would result in your ad achieving the highest placement is the one which will receive the impression."
If this was true, then it's possible that if "apples" had a higher QS than [apples] then "apples" would get the impression instead of [apples].
That is completely contrary to what AWA said. Instead of clarifying this mess.. you only threw another monkeywrench in there.
In the hypothetical scenario where both CPC and QS are the same, it should alternate between the two terms until there is enough data to make a substantial judgement on which term is 'better'. With the number of things being considered for QS, I wouldn't expect that you'd ever see two terms that have the same quality score.
AWA2
If the two terms are in the same Ad Group and share the same CPC, then which one would take the better position will come down to QS for that specific search query with the ad that is going to be displayed.
Yes, that's how it appears to work. However, is this how it *should* work? Look at it from the advertiser's perspective. If you as an advertiser are specifically targeting [blue widgets] don't you both logically think that any search on [blue widgets] would be triggering your [blue widgets] keyword, and don't you feel that Adwords should follow your bid valuation on [blue widgets] rather than your bid valuation on other broader variants (phrase match, broad match, extended broad match)?