Most advertisers are scared to death by bad traffic. They would rather turtle into their exact-match shell than see what search terms Google’s algorithm has decided are relevant. Their fears aren’t entirely misplaced. We’ve all seen some horror-shows when looking at Search Term reports. However, I would argue that Google’s propensity to show your ads for a broad (and broader… and broader-er) range of search terms is representative of just how much work the algorithm can do for you when manipulated correctly.
Why take the time?
As I wrote in the broad match post, a healthy account is going to see a certain portion of its budget spent improving the account. Which is to say, dumping a little money every month on some bad terms, non-performing ads, and other tests to find out what works, is actually optimal. If you’re not willing to take a little bad with the good then you’re probably not testing enough. Your account could be making you money, to be sure, but you couldn’t call yourself optimized.When discussing this issue with clients, I usually explain the benefits of using broad match as keyword discovery tool in two ways:
Humans (even AdWords experts) aren’t omnipotent and robots are really smart: Not only could I not sit down and think up every single remotely-relevant keyword for you but I don’t have the data to tell you which ones are better than the others. Google will think of those keywords for us, and provide us with the data to let us know just how good they are. Is “Cheapest Canoe Wholesale Disco On The Moon,” good for us? My answer is, “What does Google say?”
The keywords we didn’t consider are probably cheaper: You know who else thought of that “perfect” keyword list? Your competitors. Probably via the same experience and process that you did. The fact that Google showed you for “Cheapest Canoe Wholesale Disco On The Moon,” ten times, and it converted super cheaply one time, is now a huge competitive advantage.
We’re not talking about a huge portion of the account here, so hopefully I don’t have to sell you too hard. Testing is good. Beyond the actual data, it provides you with insights about how people actually search for your product that you can use to optimize your reach via negative keywords.
Negative match types and how to use them
When implementing negatives, most advertisers are using one match type. They’re either 100% broad match or 100% exact match. Neither is ideal, in their own ways. Here’s a brief overview of the match types and a quick methodology for their use:
Broad Match: If you can find a single word to eliminate that represents an entire category of searches, broad match negative is extremely powerful. The classic is -jobs, on “Canoe Wholesale Jobs.” Going -Canoe Wholesale Jobs would block out searches with all three included, sure, but we’re looking to take a big swing and block out any job search.
Phrase Match: Similar to the above, but much like it’s positive counterpart, the words need to be in a phrase (consecutive). This is great for when you want to be more targeted, but don’t want to block out large swathes of traffic. For example, someone who sells office desks might use “-desktop computer.” They can’t use either “desktop” or “computer” in broad, and using the term itself in broad might block out good traffic.
Exact Match: If you want the searcher to be more specific, you can use negative exact match. This is often one of the first changes we make to accounts as many people assume that broad industry searches must be good. For instance, if you do garage flooring, using [-garage] and [-flooring] might be a good idea. The searcher could be searching for literally anything related to garages, and it would be a little bold to assume you’d want to spend money for their clicks just yet.
The worst keyword match type of all, of course, is none.
Negatives: your window into search behavior
When I was in college, I had lunch with a family friend. I had wanted to ask him about the difference between getting an accounting and economics degree. He had an economics degree (so he was a little biased), but to him the difference came down to this: accountants counted up what had happened, and economists figure out what willhappen.Now, accountants are great people. My mom was one for 25 years. But the point is this: don’t be an accountant when you go through the search term report and just count up what happened over the last 30 days. Use the data that Google provides you to gain insight into searcher behavior and decide exactly what searches you’re not going to show up for and why. Extrapolate! You’ll block out more bad traffic (and less good traffic) that way anyways. Cut out big chunks when you can, little spots when you must, and let your account run free.