Campaign Analytics – Which Tactic Gets Credit?

Do you know how your marketing campaigns are performing?

It’s likely that your web analytics tool could be under or over-reporting the true performance of your advertising.

The culprit? Attribution.

Attribution – Which Tactic Gets Credit?

The easiest way to understand campaign attribution is to think about a fictional prospect, Joe.

Joe searches for “black and white sneakers” and he clicks on your paid search ad. Nothing catches his eye, but he signs up for your email newsletter.

Next week, he gets your newsletter and clicks on the “Sale sneakers” section. A pair of Onitsuka Tigers catches his eyes, so he does a little price comparison research.

Seeing that you have the best price on Shopzilla, Joe clicks your listing and buys.

Joe hit 3 different campaigns while shopping: paid search ads, email and a comparison shopping engine. So, which one gets credit for the sale?

Well, it depends.

4 Types of Attribution

Depending on your web analytics program and the options you select, your campaign analytics could be measured in one of four ways:

1. First - In this case, no matter how many campaigns the user clicks on, the sale is attributed to the first one they click. In Joe’s case, the paid search ad for “black and white sneakers” would get the credit.
2. Last - Though those other PPC or CSEs may have pitched in, here only the last tactic gets credit for the sale. In Joe’s case, Shopzilla wins.
3. Even Distribution – All of the campaigns get the same credit for participating in the sale, regardless of which one(s) had more or less influence.
4. Smart Weighting – This type of attribution is similar to even in that all tactics do (or can) share in the credit, but the distribution is uneven. Instead, it depends on the relative influence of the tactic.

The first step to understanding how attribution affects your measurement is to ask your web analytics vendor which type they use.

For example, Google Analytics attributes results to the last campaign. Knowing this reminds you that you have to be careful with your tagging.

If you use campaign tags for other elements on your site, say you want to determine the value of a particular image on your homepage without testing, you could see a drop in your results.

In the case of Google Analytics, you can use the nooverride tag to switch attribution from last to first campaigns. Remember, you must tag all of your campaigns.

Making Peace With Attribution

Now that you’re aware of the issue of attribution and know how it might be hindering the analysis of your campaigns, you have to decide what you’re going to do about it. I recommend three steps to start:

1. Address Tagging Issues – As I noted in my Google Analytics examples, the particular way your program handles campaign attribution may mean you have to change some of your campaign or site tagging so as not to throw any numbers off.

2. Pick Your Source and Stick To It – Different programs handle attribution is unique ways. Take Google’s AdWords conversion tracking, for example–it doesn’t see your other tactics, so it will take all of the credit for the sale (you can do some additional tagging to get around this).

Figure out the source that you’re business is most comfortable with and optimize from that.

3. Chart Trends from Multiple Sources – Once you’ve picked your favorite data source, you still need to validate data quality on a regular basis (quarterly or monthly).

Compare one report against the other and look for odd trends.

• Does one report show paid search going up 10% each month and another down 15%?
• Are the raw numbers within some reasonable threshold (say, 5-10%) or are they way off (20%+)?

No set of metrics will be perfect, but they should be generally going in the same direction.

How are you accounting for attribution with your campaign analysis and optimization?

Tagged as: attribution, campaign-analysis, campaign-analytics, Web Analytics