As promised, Google's Broad Match Modifiers have arrived for use in the US!
Mark your calendars, folks, this is a BIG day.
Google's graphic presents the product very clearly:
By simply adding a "+" to the beginning of any word in a keyword phrase, that word or its very close relatives becomes a required element in the user search string. BMM can be used on multiple words within a keyword phrase, and using BMM on all words in a phrase gets us back to good old broad match from 2004/early 2005.
Because we complained so loudly for so long, and because we have a reputation for using the tools they provide in smart ways, Google invited us to participate in the Beta test of this product (they did not, sadly, see fit to name it after me...GMM has a nice ring!) and we are duly excited.
There will be a temptation among some to simply use this as a cost-cutting knife. Some will simply stick a plus sign in front of each "token" and be done with it. And, by trimming out the least targeted traffic, costs will drop and will drop much more than the corresponding drop in sales.
However, we'd argue that if a broad match campaign met its efficiency targets before then making it more efficient through use of BMM misses the opportunity to maximize sales within the target efficiency metric. Simply turning the current Expanded Broad Match ("EBM") into Good-Old Broad Match will allow advertisers to raise bids and capture more of the higher quality traffic by avoiding the lower quality traffic.
Moreover, BMM will help advertisers reduce self-competition to some degree, forcing Google to serve the most relevant ad, not just the highest bid ad that's in the relevancy ballpark. This will, in turn produce higher CTR and higher conversion rates as the landing pages are more targeted to the user's search.
We advise caution with respect to doing away with EBM entirely. Some of the expanded matching is incredibly good, catching typos in model numbers and misspellings that no one could anticipate. The key is to use negatives to filter out as much of the lousy traffic as you can, and bid the appropriate amount for the traffic, given that it is less valuable than closer matches.
We'd also advise folks who have sworn off broad match entirely to experiment with BMM campaigns. Trying to catch every permutation and word ordering through exact and phrase match is impossible -- and this coming from a firm legendary for being OCD when it comes to keyword creation :-) -- so using BMM to expand the field should help.
Whether adding an extra campaign version to the layer-cake is the best approach or something less complex, testing will prove what's best for each client.
Time spent tuning these new controls will pay significant dividends.
Thanks again to the Google Team for building a brilliant product!