Watch the video below for a tutorial on putting Google AdWords conversion code on your website. For best results, watch in fullscreen or get an electron microscope.
Find all the Google AdWords For Dummies companion resources here.
Found a problem on this page? Broken link? Out of date information? Typo? Please let us know by leaving a comment below so we can fix it! Thanks!
8 thoughts on “Conversion Code”
What did I do wrong??! Were they just indecisive or was something unclear to them?
Were the clicks on different pages, or just an over-reporting error?
Where do you put the code if your “thank you” page is just a php “alert” popup notice. For instance, if tracking a lead after a prospect fills out a form and clicks a “submit” button, a popup alert displays “Thank you for your info…”. The php alert code is just a stand-alone php code file called upon clicking the button, with “” begin/end tags, and no HTML “body” tags.
Is there any reason you don’t create a separate “thank you” page? Aside from making it easy and reliable to collect conversion data, you can also move the relationship forward with new and relevant content, rather than simply returning the prospect to the page they were on before they converted.
In other words, I don’t know how to do the specific coding needed for a pop-up conversion form. Since every website is different, you may need consulting help if you want to keep the pop-up. We have a code whiz here – let us know if you want an introduction.
On page 126 under the “Specify ad delivery options”, you state: “If you set up and verified conversion tracking already, allow Google to optimize for conversions. It will save you time and money, and allow for much more aggressive testing and improvement.”
My question is, does Google really do this reliably and objectively, since their primary interest is to maximize revenue? For instance, say I want to split test two ads, letting them run for a full month using straight ad rotation (both shown evenly). The first week of the test, ad A gets 10 conversions and ad B gets only 3. Then, for the remainder of the month, ad A gets only 3 conversions per week while ad B gets 10 per week.
Since after the first week ad A is getting more conversions, will Google start optimizing to show ad A preferentially over ad B? If so, then this will mask the true results (under straight ad rotation), which would show ad B to be clearly superior at the end of the month. Wouldn’t we get more reliable data setting up our own split tests, where we know exactly what we’re measuring?
In my experience, Google does a pretty good job at this.
Yes, their directive is to maximize revenue. But half of their directive is to maximize revenue tomorrow. The way they see it, a converting customer is happier than one that doesn’t convert. So that person is more likely to use Google when they search tomorrow.
You can certainly run 50/50 (ish) tests, as long as you use different tracking URLs. Or test ads first for CTR, and then find a proven winner ad and test landing page conversion with a different system. The only danger there is two different ads attracting two different audiences, each of which would respond differently to the offer.
Besides, you don’t want to compare absolute conversion numbers. Instead, you want to find the ad that delivers the most profit per impression. So even if the rotation gets skewed, that metric isn’t effected.
Did that help?
Well, not entirely. I thought that setting up conversion tracking via AdWords would eliminate the need for different tracking URLs. It is my understanding that if doing a 50/50 ad split test for conversions, AdWords will report the results on an ad-level basis. What would the need be for separate tracking URLs?
Randall, you’re correct. If you use AdWords conversion tracking, you don’t need tracking URLs (unless you’re tracking with a system that doesn’t integrate with Google, as some folks like to do because they don’t trust the Beast with their data.
I thought you were looking for a way to avoid conversion tracking due to your dissatisfaction with their ad serving algorithm, so I was suggesting a workaround.