Broad, broader, totally unrelated…

Posted by Jan on 17 Jan 2008

Is Google trying too hard to think for advertisers?

Any search engine advertiser knows that there are very good reasons for knowing exactly what search terms users are really finding your ads for. There are quite a few ways of getting this information, which is not what this discussion is about, although I might delve into that another time… For now though, all I intend to discuss is Google’s way of trying to “help” advertisers by matching terms that Google sees as related, to the actual terms advertisers choose to bid on.

When looking at the actual phrases your ads are actually showing for, you might be very surprised and even shocked. Your ads could be appearing for terms or phrases you would never dream of bidding on and that have absolutely no relevancy to your site. Best of all is that you would never even be aware of it if you don’t check your data… As we all know, there are three types of keyword match types for advertisers to choose from, namely Broad, Phrase and Exact match. The aim of this discussion is not to teach anybody about keyword match types, but rather to delve into the “Broad” match type a little.

Before I go into my explanation of broad match types, I feel it is necessary to say something about why I even use broad type keywords: When launching a new campaign, there is absolutely no way of knowing exactly what the keywords are that people are using for your specific product. Not even the most extensive keyword research can give you all the words people in your target market are using. You can get a good indication using various available tools but in order to find out what people are really using, you have to make use of broad type keywords. Using broad types can be rewarding, but could be more damaging than anything else if you don’t know what you’re doing. Combine your gathered search phrase data with your various groups by adding relevant terms and adding irrelevant terms to your negative list and you have a winning recipe. How you do that is not to be disclosed here for now… (Most good advertisers will know exactly how to anyway!)

So, what I actually want to discuss is Google’s initiative to show your ads for terms you’re not specifically bidding on. Here’s an example of what I’m referring to: Say for instance you’ve got a website selling tennis equipment. On your site you have a tennis shoe section, for which you have created a specific adgroup with specific keywords and specific ads that go to specific relevant landing pages. You might have the term “tennis shoes” as a broad match type in your keyword list as it is absolutely related to your site and specifically this section. By having this key term in your list, you would also be able to pick up and gather new terms related to tennis shoes, which will help you build and expand your keyword list. The reason I say that you should check your data every day is because even if your ads aren’t supposed to, it might also show up for searches like “sport shoes” or anything to do with sport and shoes… In some cases this could be very helpful, which is obviously the reason Google is doing it, but in other cases, it can destroy your budget if you don’t pick it up. I know that very few advertisers even use negative match types, which makes me wonder, just how many millions of dollars are spent daily on clicks by totally unqualified traffic to websites across the internet. This Google technology is called “Expanded Keyword Matching” and is only intended to show your ads for searches that are still relevant to your terms, but as I’ve been saying, that is not always the case as what Google sees as relevant to your site, is not always the same as what you might see as relevant. On the other hand, it is also this very technology that helps advertisers who know how to use their data, build and expand their keyword lists.

So how do we use this to our advantage?
By simply using the data at our disposal to regularly add relevant terms to our keyword lists and even more regularly add irrelevant terms to our negative match keyword lists. Remember to make sure you know exactly how negative matching works before you start blocking traffic to your site though…