Finding invaluable and useless negative match keywords
Aug 1, 2008 Google Adwords, Google Analytics
Anyone who’s ever started a campaign using broad match keywords knows the pain of generating useless clicks on irrelevant fringe keyword combinations. For instance, if you’re targeting a broad match keyword combination like “used car” you don’t necessarily want to get clicks from people searching for “used car salesmen suck”. But how much is too much when it comes to negative keywords?
In the aforementioned example, we’ve found one of several keywords we’d probably add (instinctively) to a negative match keyword list in a Google Adwords campaign. If we were only interested in those people that are really looking to buy used cars, we probably wouldn’t think twice about excluding a generous amount of keywords that are unlikely to convert - at all. Many industry insiders have shared default negative match keywords they use exclusively on ALL their campaigns, such as:
- myspace
- aol
- sucks
- xxx
- www
Although it does make sense to exclude these by default across several similar campaigns, it doesn’t necessarily help campaigns that make use of exact match or longer tail phrase match keyword combinations. Common sense dictates you should only really exclude those keywords that hurt your campaigns, for instance, fringe words that will sap expensive clicks from poorly optimized and expensive “bread and butter” keywords.
Be selective when choosing negative match keywords, especially if those negative match keywords may possibly come back bite to bite you later.
A classic example of what I term “negative match revenge” occurs when retailers add products to their online stores that they didn’t have in stock or prior access to in the past. My bet is that when the iPhone came out, many cellular retailers that didn’t have iPhones or compatible carrier partnerships would exclude the term from their campaigns. Now that the iPhone is more readily available across different cellular bands, many of those same retailers may suffer “negative match revenge”.
The problem arises from having keywords in both your negative match keyword list and your targeted keyword list. Basically the two instances cancel each other out. Although my “iPhone” example is plainly simple and somewhat obvious, there are many more opportunities for “negative match revenge” across subtle keyword selections, so be wary.
Reports of advertisers keeping lists of 10,000+ negative match keywords seems unnecessarily excessive, except for very poorly optimized campaigns. What do I mean by poorly optimized?
- Campaigns set to “All regions” rather than specific countries (at the very minimum)
- Campaigns spanning multiple languages
- Extreme use of broad match keywords
- Campaigns set to both search, partner search, and content network placements (sigh…)
- Ad groups with tons of unrelated keywords
For advertisers that use one or any combination of these tactics, they really do REQUIRE negative match keywords lists in excess of 1,000 keywords long. But I’m guessing what they really require is a better search marketer…
Most advertisers are now aware that Google provides tools such as the Search Query report, which gives a breakdown of those keyword combinations that trigger ads on the search network. However, the Search Query report often spits out useless match information such as “657 other unique queries” which can leave advertisers in the dark. Word Tracker, a subscription based keyword research tool can help in these instances by offering advertisers performance indices.
On the other hand, advertisers with a little ingenuity can slim down those numerous unique queries by adding common-sense keywords to their campaigns. The more you optimize your keyword lists, the more information a Search Query report will give you, so keep plugging away at it.

















August 6th, 2008 at 2:36 pm
[...] We all know the pains associated with broad match keywords that may have no relation to what you are actually targeting in your PPC campaigns. The folks over at PPC-Advice.com offer up some useful strategies to counter this issue. We typically employ negative keywords as the weapon against these broad match keywords. This post outlines what negative keywords are useful and what aren’t. [...]