Google AdWords for Mortgage Brokers Are You Making Basic Mistakes? by Dennis Yu

by 09 Mar 2010

Pay Per Click (PPC) advertising in the mortgage industry is confusing:

  • There are 124,000,000 searches for mortgage related terms on Google per month in the United States, according to The average cost per click is $13.69—but terms like “mortgage refinance” are up to $32 per click and “loans” can be $3 a click, but are more plentiful. What words should you use?
  • The bulk of the traffic is among 96 national advertisers with deep pockets—you already know who they are. Should you try to compete against them?
  • Some leads are worth more than other, depending on factors such as type of loan, value of home, credit score, and so forth. How do you obtain the right traffic at the right price?
  • What about the multiple search engines and display ad networks—is it worth going beyond Google AdWords?
  • Do you hire someone else to do this or keep in-house—and, either way, how do you make your PPC accountable from a ROI standpoint?
In this month’s column, we’ll start with keyword strategy—to discuss how to organize your keywords in a way that creates the highest relevancy and Quality Scores. The net result is that your cost of traffic will decrease, your conversion rates will increase, and your campaigns will be easier to manage. In subsequent months we will cover bid optimization, landing page optimization, search vs. content networks, analytics, and call tracking.


1) Broad matching on most of your traffic:

If you already have a Google AdWords campaign, odds are that you have a few keywords in a few ad groups that are generating the bulk of your traffic. You might even have 2 million keywords, as we have seen from auditing the portfolios of large financial services companies - but the odds are that just a dozen keywords make up the lion’s share of the cost.
So look at those keywords, sorted by cost descending in Google AdWords Editor or your favorite tool. I will bet that most of those terms are on broad match, meaning that your term can loosely match for any search containing that word. So even though you think you’re bidding on “mortgage,” you might actually be buying “mortgage rate calculator,” “what are mortgage prices in London,” or any number of queries.
Below is a screenshot of a typical Google search on “mortgage.” You can see how Google uses their “suggest” tool to guess at what you might be looking for.
The solution is to create exact match copies in their own ad groups, then place a large number of negative keywords in your broad matched ad groups. If that sounds like a bunch of jargon, let me explain. Next to each keyword, you can specify match type. If you are buying “mortgage” on an exact match, you will create a far higher quality click, since it will not match against hundreds of thousands of searches that you did not intend.
At the same time, you still do want broad match traffic, since it gives you a wider traffic base. You just have to be smart about it and use negative keywords (negatives) to exclude traffic you do not want. For example, you might add “bankruptcy” and “bad credit” as negative keywords, so that if those words are in a user’s query, your ad won’t trigger.
The broader the keyword, the stronger the negatives have to be to protect you. For example, if you bought the keyword “loan” on broad match, you’d have to negative out “car,” “boat,” “automobile,” “student” and other types of searches that would contain the word “loan”. In this example, it’s probably foolhardy to bid on the term “loan,” since there are too many other words to try to exclude.
You certainly don’t want to accidentally be buying “personal loans Australia” or any related terms. You may be surprised how much international traffic there is in mortgage lead generation, as well as searches that are not real estate related.
In general, you should aim to have at least a few hundred negative keywords at the campaign level and at least as many at the ad group level. The quantity depends on how broad your keywords are. If you are bidding on “mortgage refinance rates” on broad match, you will not need as many negative key words as “mortgage” on broad match by itself. You’re allowed to have up to 10,000 negative keywords, so you might as well use them. 
Need help generating negatives? You can run a Search Query report in AdWords to tell what searches actually triggered your ads. Or just use data from your Google Analytics or third party keyword research tools to determine what should be a negative. I am personally a big fan of using Google Suggest and Google Wonder Wheel—after all, Google will have better data on Google than anyone else; it would stand to reason.
2) Dumping all your keywords into the same ad group or campaign:
Closely related to mistake #1 mentioned above, most PPC advertisers will jam all their keywords into a single ad group, even though each keyword has a different intent.
Think about it. Someone searching on “mortgage calculator” is expecting to see something quite different than “Denver mortage broker.” Thus, you should have a different message for each of these searchers. 
An ad group is about grouping ads, not grouping keywords—otherwise it would be called a keyword group. Think of an ad as a message that speaks to a particular intent. It could be whether mortgage rates are at their lowest, how much house someone can afford or maybe they want to comparison shop at a or If they are performing comparison shopping, why might they want to talk to you versus going on-line?
There are two schools of thought in keyword grouping—one is the brute force approach and the other is the hand-made approach. Neither is better than the other, as both approaches are needed at different points in the campaign building and optimization process. What they have in common is they take a large bucket of keywords and break them down into smaller, tighter groups where words that are within an ad group will have approximately the same meaning.
We’ve started from a large bucket of keywords—you can choose your favorite keyword research tools or just manually dump in a bunch of keywords. And then we begin to divide up the keywords into groups, filtering down further and further until we have a nice handful of ad groups. Don’t go crazy here—you might want to start with just a couple dozen ad groups, since if you make 200 ad groups off the bat, they will become unmanageable and you’ll likely exceed your account limits anyway.
Branded traffic is separated out because they perform differently. A brand search is someone typing in “lending tree,”as “” or another search where it is clear what they want. Those searches are of high volume and also convert well if you have the right ad to lead them your way. Bidding on branded terms is topic for another article—but suffice it to say that Google says it’s okay and that most companies are doing it.
When you separate out your keywords into tightly knit groups, you’ll see several things happen:
  • CTR (Click Through Rate) increases, since you’re now showing an ad that is relevant to the user’s search query.
  • Google Quality Score increases- CTR is the largest component in the Quality Score metric, which has over 100 variables. We’ll talk about Quality Score (QS) in another article, since it’s so important.
  • Average CPC decreases relative to average position—in other words, you’re now paying less for traffic. Think of Quality Score as a discount factor. If you have a QS6 and someone else is QS3, because your QS is twice theirs, you will pay half as much per click, all else equal. You should shoot to be at least QS7 on all your keywords. If not, you have a problem—likely one of the ones that we have mentioned earlier.
There you have it! Broad matching on most of your keywords and not organizing your ad groups into tightly-knit themes are what kill most PPC advertisers. If perhaps you haven’t made a PPC campaign yet, please contact and we can provide you with a free templated campaign you can start with—no strings. This is a courtesy to all subscribers of The Niche Report. If perhaps you have a campaign that does not have strong negative keywords, we would also be happy to provide you with a basic list of negatives that you can customize.
Next month, I will talk about bid optimization—to discuss how much you should be willing to pay for each keyword.
Dennis Yu is the CEO of BlitzLocal, an online ad agency specializing in local search and social media—helping your business get found online and driving calls to your phones. Clients include Equifax Corporation, California Pizza Kitchen, World Wrestling Entertainment, Quiznos, March of Dimes, and businesses of all sizes that want to drive more leads online. Dennis is a noted international speaker and author, having appeared on NPR, CBS Evening News, TechCrunch, KTLA, Search Marketing Strategies, and other media.



Is TILA-RESPA a good or bad thing long term?