Modified broad match vs phrase match strategy - Google Adwords

[dsinger]

Hi All,

I am looking through a client account that is very mature (10+ years running) on Google AdWords. As soon as it became available, this client adopted a modified broad match (MBM) strategy and has removed all phrase match and exact match keyword types. The account has hundreds of thousands of active keywords. Over the past few years, the CPCs have been rising. While I know that market values of keywords in general have risen consistently year after year, I speculate that this client is actually causing their own prices to go up faster than they should. I have a couple of questions regarding strategy that I am considering that I want to know if anyone else has any experience with...

1. by having many MBM versions of the same keyword, is it possible for cannibalization to occur for most of the variations?

Example query: new red running shoes
variations Ad group 1: +red running shoes, +red +running shoes, red +running +shoes, red running +shoes
variations ad group 2: +blue running shoes, +blue +running shoes, blue +running +shoes, blue running +shoes

based on the logic of MBM, the possible matches to this query from the available variations are +red running shoes, +red +running shoes, red +running +shoes, red running +shoes, blue +running +shoes, blue running +shoes. So, if the performance of those blue variations trump the more closely related red variations, this searcher may actually see an ad about blue running shoes, even though they have indicated they are more interested in red.

2. in terms of cost, I would anticipate that MBM keywords are more expensive than their phrase match counterparts. can anyone confirm or deny this?

My thoughts are that with several years of actual search terms being collected, this client should be able to do a great job of covering almost every variation of keyword that customers have used and create a strong list of phrase match keywords to satisfy all relevant queries. MBM keywords seem like they are a lazy way of getting traffic at a higher cost that can actually cannibalize close variations that exist in the account, causing the wrong ad to be shown based on matching/relevancy and a higher price CPC in the long run.

Thoughts?

[flowsimple]

Your answers are inaccurate.

1. A higher bid does NOT always win. It's the QS X the bid. If a keyword with a high QS X low bid is greater than a keyword with a low QS x high bid, then the lower bidding keyword can still trigger the ad.

2. There is NO pattern to MBM keywords being more expensive than phrase keywords. The opposite can often be true for high intent keywords where it has been discovered to result in more conversions or lower CPA.

[flowsimple]

This is an interesting strategy question. Here are some answers + my thoughts.

1. Cannibalization. (Cannibalization is sort of the wrong word, because you do go on to talk about the blue ad group "trumping" the red ad group.) Assuming both keywords have equal bids, Google will always trigger the ad of the associated keyword with the better quality score. (If there are different bids, then simply multiply the bid x QS to determine which would ad would trigger.) The overlap/trumping can still have an adverse effect, whereby the ads are not always relevant to the search intent, thereby reducing CTR and then adversely affecting the Quality Score, and the cycle continues. The reduced Quality Score is the likely reason for the increased average Cost Per Click.
2. **Actual cost of different match types. **Is MBM or broad match always more expensive than phrase and exact. No. Sometimes exact and phrase are more expensive than their broad counterparts because high intent keywords come at a premium. Also, MBM covers the phrase match keywords, and one needs to remember that the CPC column you see is just an average of all the keyword clicks combined. You can see the true CPC by looking at the Detailed Search Query report.

Hope that helps!

[JasmineA]

Hey DJ!

1. Yes.

(i have to say, I really hope that your client/company isn't focusing on a MBM with something as broad as : +red blah blah, because that's nearly as bad as just broad on the term Red, ok, moving on to answer your question)

Google has gotten significantly better at making the decision to show a Red ad v a Blue ad when the searcher really wants Red, however if the bids are higher on Blue then yes, blue will win out. The best way to protect yourself against bid wars is to build up your negative keyword army (yes, all of them. yes, this is time consuming and it blows a little).

Also, these three different types of MBM should NOT be in the same adgroup. +Red running shoes & red +running +shoes are two completely different types of requests. For instance, think of MBM as phrase match. I wouldn't put "red" and "running shoes" and "shoes" in the same adgroup (at least, that's my opinion).  But see the above point, you'll need an ever growing & large list of negative keywords to keep this under wraps, which is a second full time job.

2. I haven't looked at it recently, but MBM should definitely get a bit more costly than phrase. Exact being least expensive, or at least on par with Phrase in my experience.

But the bids & strategy are really the major contributing factors. I typically stack my bids highest for exact, then phase, then MBM & broad at the lowest. So I refuse to let them get high. But also stacking bids helps to prevent that cannibalization that you might be seeing.

Hope that helps!