Exact match when using phrase match?

Hi there,

I have some AdWords campaigns where I use both phrase match and exact match. I used to have them in separate campaigns, but it didn't seem to make a difference, plus I thought that I might be competing against myself.
So no I have campaigns with (example) [blue widgets] and "blue widgets" in the same campaign.
I was then wondering if having the exact match make any sense at all since it's already covered by the phrase match, so it just muddies up the campaign overview.

Any thoughts on this?

Thanks,
Thomas

Bump

thomaste said:

Hi there,

I have some AdWords campaigns where I use both phrase match and exact match. I used to have them in separate campaigns, but it didn't seem to make a difference, plus I thought that I might be competing against myself.
So no I have campaigns with (example) [blue widgets] and "blue widgets" in the same campaign.
I was then wondering if having the exact match make any sense at all since it's already covered by the phrase match, so it just muddies up the campaign overview.

Any thoughts on this?

Thanks,
Thomas

Click to expand...

unfortunately, i don't think a single cookie cutter answer for this. It all depends!! it depends on:

1. what are you ranked for "blue widgets"
2. are your competitors doing ad-words
3. customer intention when they search for exact match and phrase match of your keyword

initially, I used to have both exact match and phrase match in a single campaign. from my google search, i figured that those who did the exact match search, were the type of customers who came to my site with the intention to purchase. those who used the keyword in a phrase, did not covert that often. because of this i changed my campaign to only show my ad-word for exact match type.

hope this makes sense

You're not “competing” against yourself, no matter how you structure your campaign. The system simply chooses the most appropriate keyword based on the search term.

It first checks if there's an exact match (and now close variants). If none, it checks if there's a phrase match. If no phrase match, it moves on to broad match, the modified version first and finally the non-modified where it can really go broad. I don't recommend using the non-modified broad match.

On to the root of your question. As you point out, if you didn't have the exact match, ads can be triggered on the phrase match. So why have the exact match?

For one, you can gather valuable data on how people search. Do they tend to search [blue widgets] more often than “blue widgets”? You can see that very easily when they are in the same ad group. You could of course answer that question using the search term report, just a bit more involved.

Also, while I have never seen anything official, I suspect that exact matches are given priority over other matches. Suppose during an auction there are five advertisers using the exact match and one who doesn't. The system may not show the ads of the one using just phrase match as often. There are other factors of course, mainly your ad ranking. But you may find having only phrase match reduces your impression share and maybe even your average position.

In short, my answer to your question is to use all match types to cover all bases. From a data analysis viewpoint, it can help determine trends which you can act on to improve the campaign.

Thanks so much for taking the time to reply. Really appreciate it, and it makes very good sense what you say.
So until further, I'll keep both.
I'm up against a lot of competition, and have a limited budget, so I've dropped 'Broad' since I needed to focus my budget on the most narrow group of potential clients.

Thanks and have a nice weekend

thomaste said:

Hi there,

I have some AdWords campaigns where I use both phrase match and exact match. I used to have them in separate campaigns, but it didn't seem to make a difference, plus I thought that I might be competing against myself.
So no I have campaigns with (example) [blue widgets] and "blue widgets" in the same campaign.
I was then wondering if having the exact match make any sense at all since it's already covered by the phrase match, so it just muddies up the campaign overview.

Any thoughts on this?

Thanks,
Thomas

Click to expand...

This is a very simple easy answer: You only need the PHRASE match of any keyword, and Google has actually confirmed this in many recent postings. So having "blue widgets" is much better than [blue widgets]

Thanks for confirming this. However, I believe that [blue widgets] would be the phrase match, correct?

Thanks,
Thomas

Sorry, you're of course right:
[blue widgets] = exact
"blue widgets" = phrase.

Thanks again

The adRank in auction time determines which keyword to match the search query. That is why to set higher bids for the exact match and lower the bids for phrase and broad match types.

One way to structure the account is to create different adgroups for different match types and to use negative keywords. For example if you have 3 adgroups for 3 match types, you need to add the keyword as the exact match negative keyword in your phrase and modified braod match adgroups. This way you can control which ad/adgroup will be served.

persiaseo said:

The adRank in auction time determines which keyword to match the search query. That is why to set higher bids for the exact match and lower the bids for phrase and broad match types.

Click to expand...

Nope. Which keyword the system uses has nothing to do with the keyword's ad rank. Ad rank (which is QS times bid for those who don't know) only applies in the step to determine how to rank ads that are eligible, not in the selection of which match type to use as explained earlier. In fact, internally in regards to QS, the system treats them all the same. The QS for a keyword in broad (not BMM), phrase and exact are always the same. So having different bids on these will not affect which keyword actually gets used. It will of course affect your ranking.

persiaseo said:

One way to structure the account is to create different adgroups for different match types and to use negative keywords. For example if you have 3 adgroups for 3 match types, you need to add the keyword as the exact match negative keyword in your phrase and modified braod match adgroups. This way you can control which ad/adgroup will be served.

Click to expand...

This part of your comment also doesn't apply as per my response above. You cannot control which match type the system uses.

LucidMarketing said:

Nope. Which keyword the system uses has nothing to do with the keyword's ad rank. Ad rank (which is QS times bid for those who don't know) only applies in the step to determine how to rank ads that are eligible, not in the selection of which match type to use as explained earlier. In fact, internally in regards to QS, the system treats them all the same. The QS for a keyword in broad (not BMM), phrase and exact are always the same. So having different bids on these will not affect which keyword actually gets used. It will of course affect your ranking.

Click to expand...

Yes the quality score for all the match types are the same but as I said earlier you can give set the bids higher for exact match to raise the adRank for the exact match type so you can control which matching options can trigger the ad.

LucidMarketing said:

This part of your comment also doesn't apply as per my response above. You cannot control which match type the system uses.

Click to expand...

If you structure your account based on match type by creating separate adgroups for different matching options you can control which match type to trigger the ad by utilizing the negative keywords.

persiaseo said:

Yes the quality score for all the match types are the same but as I said earlier you can give set the bids higher for exact match to raise the adRank for the exact match type so you can control which matching options can trigger the ad.

Click to expand...

The issue is that the system in the first pass, the first step, does NOT choose the keyword based on adrank. That happens in other steps and the step that determines the keyword occurs before the ranking. Therefore, you cannot control which keyword is used. You are also forgetting that adrank is not an exclusive property of the keyword, it is really the kw-ad combo. So if you have two ads, it's the adrank of that kw+ad. You are looking at the keyword in a vacuum which it certainly is not.

https://support.google.com/google-ads/answer/7502501?hl=en

Key sentence in that link: "the preferences below are used to determine which keyword is used to enter an ad into an auction". In other words, that's the first step, which keyword best matches the search term. No mention of QS or adrank.

It's only in the third step that QS is taken into account. They don't mention multiple ads but I guess they wanted to keep it simple. QS is affected by the ad much more than the keyword.

And as they show below, it's not set in stone otherwise the system wouldn't really work very well and it would become more of a straight best-bid-wins auction. So there are some exceptions and system decisions just to mix things up and gather useful data.

LucidMarketing said:

https://support.google.com/google-ads/answer/7502501?hl=en

Key sentence in that link: "the preferences below are used to determine which keyword is used to enter an ad into an auction". In other words, that's the first step, which keyword best matches the search term. No mention of QS or adrank.

It's only in the third step that QS is taken into account. They don't mention multiple ads but I guess they wanted to keep it simple. QS is affected by the ad much more than the keyword.

And as they show below, it's not set in stone otherwise the system wouldn't really work very well and it would become more of a straight best-bid-wins auction. So there are some exceptions and system decisions just to mix things up and gather useful data.

Click to expand...

Thank you for the informative post and the reference you provided.

Consider this scenario:

Adgroup 1
Keyword: [blue widget]

Adgroup 2
Keyword: "blue widget"
Negative: [blue widget]

Adgroup 3
Keyword: +blue +widget
Negative: [blue widget], "blue widget".

This is the structure I am talking about if you separate them by match type.

It doesn't matter how you structure your campaign, whether one group with all match types or a match type per group.

Say the search term is "round plastic blue widget".

The first step is to see if you have any keywords that are eligible. The order of preference is exact then phrase then BMM then broad.

You don't have an exact match and the best option is the phrase "blue widget". This regardless of your bid or your ad quality.

You are now eligible to have an ad shown. The second step is therefore to calculate an ad rank for all eligible advertisers. Just before that however it would need to decide on which ad to show, assuming you have more than one active ad, because it needs to know the ad's quality to calculate the ad rank.

The ad rank is used only to rank all eligible advertisers. Its name says it, there's no ambiguity. Think about it: if you are right, then the system would always choose the keyword with the higher ad rank without regard to how well it matches the search term. In other words, if the phrase match had the higher ad rank (which you are saying compared to your own ads and keywords, not competing advertisers), it would select the phrase match over the exact match even if the search term exactly matches. That is clearly not the case.

I've seen this said many times here and elsewhere that "your highest ad rank keyword is chosen". I don't know how it got started but it simply is not true and it can easily be proven.

I exclusively use phrase match. It is very important to check your 'search terms' often and set the right negative keywords accordingly.