Adwords, like any tool, is a double edged sword. If used improperly, it can flush your money down the drain. But used correctly, it can be a very powerful driver of sales to your online store. The main problem however, is that most new users fall into the money wasting category.

One of the biggest mistakes that newbie Adwords users make is to misuse broad matching keywords with their campaigns. To illustrate my point, I’d like to tell you about some of the boneheaded mistakes that I made early on with our Adwords campaigns.

As I mentioned in some of my previous articles on How To Use Adwords The Right Way With Your Online Store and The Realities Of Using Adwords Effectively, finding the right keywords to use is an iterative process that requires patience and the right tools. And the only way to refine your campaigns properly is by refining your analytics tracking as well.

Broad Match Can Waste You Money

The funny thing about search is that there are always people out there that perform searches that you would never expect. For example, we sell linen napkins in our online store so early on in our Adwords campaign, we decided to use the keywords “linen napkins” (even though I’m using quotes here, assume that I’m using broad match) with our Adwords campaigns. Back when I was more naive, I used to think that this was good enough. But let’s take a look at some of the searches that resulted in clicks to our site.

• Vintage Linen Napkins
• Green Color Linen Napkins Cheap
• Shamrock Linen Napkins
• Irregular Linen Napkins
• Essance Linen Dinner Napkins
• Used Linen Napkins Wholesale

Looking at the list above, we don’t sell any of these things so we are basically throwing our money away for these search clicks! The proper way to refine these campaigns would have been to do the following…

• Add negative keywords. ie. Don’t match on “vintage”, “green”,”Irregular”,”Essance”,”Used”
• Add additional keywords or use exact match. ie. Match only on “white linen napkins” which is the only color we sell.

Using Google Analytics The Right Way

Believe it or not, Google Analytics does not provide you with the exact search query information by default. Instead, it just shows you the number of people that clicked on your campaigns based on your campaign keywords only! Just to be clear, all of the above search queries would have been categorized under the keyword “Linen Napkins” by default which basically tells you squat.

The right way to use Google Analytics is to create your own filters to allow you to see the actual search query that was used to land on your webpage. But to do that, you need a tiny bit of technical know how which I will teach you.

Anatomy Of A Search Referral URL

Before we begin, it’s important to understand what a referral URL looks like coming from a search engine. Below is an example from a Google search.

http://www.google.com/search?hl=en&client=firefox-a&hs=8aw&rls=org.mozilla:en-US:official&q=getting+good+grades&start=40&sa=N


As you can see, there is a lot of information embedded in the URL and we want to be able to extract the search query information in our analytics reports so we can correlate our adwords campaigns to actual searches.

The field that is the most important in the example URL above is the &q=getting+good+grades. The &q field indicates the actual query that was used was “getting good grades” and is the information that we want to extract out.

Create A Google Analytics Filter

What is an Analytics filter? Filters are applied to the information coming into your account to manipulate the final data in order to provide accurate reports. In this case, we want to filter out the search query so we can correlate it with our Adwords campaign keywords. Please note that the way the following example is written will only work for Google searches. I did it this way to make the explanations simpler.

Here’s a step by step on how to do this

1. On your analytics home page, create a new profile and name it “Adwords Queries” or whatever you want
2. Click on the “edit” button next to your newly created profile
3. Scroll down and click on the “Add New Filter” button
4. Name your filter “Adwords Query” and fill out the rest of the information exactly as specified below.
   Filter Type:Advanced
   FieldA->ExtractA = Referral &q=([^&]*)
   FieldB->ExtractB = Campaign Medium (ppc|cpc)
   Output To->Constructor = Custom Field 1 $A1
   Hit “Save Changes”
5. Create another filter, name it “Campaign Term To Query” and fill out the rest of the information as specified below
   Filter Type:Advanced
   FieldA->ExtractA = Custom Field1 (.*)
   FieldB->ExtractB = Campaign Term (.*)
   Output To->Constructor = Campaign Term $B1 ($A1)
   Hit “Save Changes”
6. On the Profile Settings Screen, reorder the filters such that the “Adwords Queries” filter is applied before the “Campaign Term To Query” filter
7. You are done!

OK…What The Heck Did We Just Do?

Essentially, the first filter is used to extract out the interesting portions of the referral URL string. For example, let’s break down the following line.

FieldA->ExtractA = Referral &q=([^&]*)

It’s cryptic, but basically it’s parsing the string looking for an &q= symbol. Then it’s grabbing everything that comes after it until the next & symbol (ie. the search query) and shoving it into a variable called $A1. $A1 is used to denote that we want the 1st parenthesized value in field A.

FieldB of this filter merely limits the pattern match to CPC and PPC campaigns only and creates a new custom field.
FieldB->ExtractB = Campaign Medium (ppc|cpc)
Output To->Constructor = Custom Field 1 $A1

The second filter doesn’t do much. It merely correlates the search query with your google adwords keyword so that it shows up in your analytics report as shown below.

cloth napkins (blue+gingham+cloth+napkins)

Voila!

I apologize if this post was a bit technical but sometimes you have to tap into a tiny bit of code in order to get what you want. Once you have these reports going in analytics, you should have no problems refining your adwords campaigns! Be sure to use negative keywords in your campaigns as soon as you discover a false search.