How the AdWords Attribution Tool works

Why do my AdWords Conversions Do not match those that shows me Analytics? Understanding The attribution models is essential to be able to analyze the results correctly and take the appropriate measures. Each tool or platform we use uses a attribution model Different and if we are not aware, we can be conducting a distorted analysis that leads us to errors.

What happens when a user clicks on an AdWords ad and then on one of Facebook before converting? Who “takes” the conversion?

These questions have their answer in the attribution model itself in AdWords or Facebook, or in the Analytics if we rely on this tool to analyze the metrics.

Today we will see how the attribution model works in AdWords. We’ll leave for the attribution model another post on Facebook and all that are available in Analytics, you have to read them all to understand how to analyze the data correctly and become a real SEM Manager ????

AdWords attribution study

Within AdWords we can analyze how campaigns, ad groups or keywords contribute to other campaigns, ad groups and keywords of the account. In addition, we will see how long the route to conversion lasts and how many interactions with the campaigns are produced to meet the target.

To see all this, navigate through the main menu to access “Tools” and select «Attribution»:

Here we find an overview of the conversion actions, their value (if assigned to it), the time that passes until the user converts and the clicks and impressions that occur on average before converting.

Now we are browsing through the menu on the left:

Major conversions

Displays the history window menu to see the difference in the number of conversions depending on the window you have chosen.

Assisted conversions

In this screen we can see how campaigns, ad groups, keywords or even “concordance types” helped to convert to other campaigns, ad groups, keywords or “concordance types”.

In this case, campaign 1 and campaign 2 contributed to another campaign having a conversion respectively. That is to say, the user clicked on campaign 1 and ended up converting after clicking on another campaign other than 1. And the same with campaign 2.

The last indicator, post-click conversions of the last click is an indicator based on a campaign, ad group, keyword, or concordance type in the path to the conversion. Closer to zero would indicate that it closed more conversions than it helped to close. If it is greater than 1 it would indicate that the parsed element is helping to close conversions rather than completing conversions on its own.

This report is very valuable, because it allows us to distinguish keywords and campaigns that although seem to have no results contribute significantly to the total conversions. Generally we find ourselves here with the generic campaigns, that serve to capture the user and later the user will convert when he performs a more specific search.

First click Analysis – Last click analysis

Displays conversion data Depending on the campaign, ad group, keyword, or match type in which the user first clicked or was last before converting.

Main Routes

It teaches the most frequent ways towards conversion. In the drop-down we can choose at what level we want to see it: campaign, ad group, keyword.

Time Lapse

This screen is very interesting to corroborate that the conversion window that we apply in our campaigns is the ideal one. It tells us the time it takes since you click on an ad until the conversion occurs.

Route Length

How many clicks occur until the user converts. It will help You know if your product/service requires a lot of interactions with the campaigns to the conversion: price comparison, study of characteristics, etc.

Attribution Models

This report shows a comparison of the models of attribution in AdWords, between our own campaigns. We will also see how the use of one model or another influences our cost/conversion or registry of conversions (percentage of change):

• Last click: assigns the conversion to the last keyword the user clicked before converting.
• First click — assigns the conversion to the first keyword the user clicks.
• Linear — assigns the conversion uniformly between the clicks of the path to the conversion.
• Temporary depreciation: you will have more weight clicks that are closest to the conversion.
• Depending on the position: the first and last click Obtain 40% of the conversion merit. The other 20% is evenly dealt between the clicks of the route.

The AdWords attribution model as a channel

The attribution model that AdWords uses as a channel is known as the “last click of AdWords”, ie, Adwords will register a conversion whenever the user who converts has clicked on an ad within the conversion window Specified.

No matter what happens next, the user can click on an Ad and not convert that day. You can see an ad on Facebook the next day, click and Convert. In this case AdWords will continue to register a conversion, because the path to the conversion of that user was clicked on one of his ads. We will have 2 conversions in our data (1 of AdWords and 1 of Facebook) but really only one will have produced 1 conversion.

This is the usual problem when the advertiser uses more than one advertising platform: Facebook, AdWords, Twitter… If 2 or more channels are involved on the way to the conversion, to which one is assigned?

The easiest solution we have to our hand is Google analytics. The Analytics measurement model is “Last indirect click”, which means that Analytics assigns the conversion to the last channel that the user clicked before converting, as long as it is not direct traffic. Therefore, it gives a more realistic view of the contributions of each channel as far as conversion is concerned. As this topic requires a greater depth, we will analyze this problem and its solutions in our next post. Will you join us?