Multiple Match Type Smackdown
In Geekfight, a Razorfish expert challenges the published wisdom of another search pundit. Today, Razorfish Group Director Adam Heimlich takes issue with a recent blog post by Charli Rogers, UK Client Services Director for Efficient Frontier.
Rogers’ post “Strike the Right Match Type” has a lot of good search-geek information on multiple match type buying. We wholeheartedly concur with her main point: Search marketers should avoid over-reliance on broad match, and, whenever broad match is used, run query reports regularly to find new negatives and minimize irrelevant matches. Razorfish search clients benefit from this best practice.
But in Part 2 of her post, Rogers gets into the crucial “how to” of query report expansion, and gives a bit of advice that rips the “best” right out the practice!
She says not to triple-match all keywords – sensible enough given how big that can make an account – and explains as follows:
“As a rule of thumb, consider how many words make up the keyword. If there are more than three, there is no need to add exact/phrase match versions to begin with. Use your common sense to inform what does/doesn’t need to be triple matched; cheap flights is a high volume generic which needs to be on all match types, but find a three star hotel in Paris probably only needs to go on broad match in the beginning…”
Rogers seems to be saying that buying long phrases on broad match alone doesn’t cause the same loss of advertiser control that we know occurs with short phrases on broad match alone. She implies that “find a three star hotel in Paris” won’t broad match to a high-volume query like “Paris hotel” or even “Paris.” Readers of the Efficient Frontier blog who took its advice can learn quickly from query reports that they’ve been misled.
In fact, token length doesn’t mitigate the risk of unwanted broad matching. Advertisers who experimented but later chose not to bid at all on “Paris” and “Paris hotel” would be most in danger of broad matching into those extremely expensive auctions. The matching algorithm is designed to fill up SERP inventory! If an advertiser with established relevance wants the power to opt out at will, all his broad match keywords must be reigned in, regardless of their length.
The smarter strategy is to build on exact match and either double up on broad, or, if account size is a factor, restrict broad match to short phrases. You don’t actually need long phrases broad matched: “Paris hotel” alone will broad match to all relevant variations omitted from the account. Adding these ON EXACT through regular query-report expansions maximizes advertiser control.
Confusion about match-type strategy is widespread in the industry. Until recently, even Google advised against the most efficient method. We thank Charli Rogers and Efficient Frontier for raising the issue, and cordially invite their response.
Have an idea for GeekFight! Email us at razorfishsearch@razorfish.com






You’re right; in isolation, it is naive to assume that the number of words in a keyword is an accurate measure of how easily controlled they will be. The recommendations in my post were to consider all of the practical steps outlined, when working with match types, including adding and triple matching high-volume head terms (which should be revisited on a very regular basis through the use of search query data). Perhaps, instead of using the example cheap flights I should have used hotel Paris, or Paris hotel, to better make it clear what is best practice; all campaigns should have a mixture of head, mid and tail terms.
Google’s matching algorithms are, as you rightly point out, a means for them to better fulfil more user searches. The onus is on us, as search marketers, to understand how to best control how our keywords are being served.