Adobe Analytics and Search Engine Data - dummies

Adobe Analytics and Search Engine Data

By David Karlins

One key advantage of analyzing data with Adobe Analytics is to help drive your marketing and advertising strategies. Once you dive into Adobe Analytics, you’ll see how the platform can be used to tie the data from search engines into your marketing efforts.

A key advertising channel for all brands occurs on search engines such as Google, Bing, and Yahoo! Companies apply two types of tactics to increase the visibility of their brand on search engines: search engine optimization (SEO) and search engine marketing (SEM, or paid search).

Analysts need to analyze behavior coming from search engines as a channel as well as distinguish between paid and natural. The data helps them determine how the channel affects behavior and conversion rate.

Adobe Analytics collects data in several search-focused dimensions, but they are unfortunately less reliable than the marketing channel and referrer dimensions. Our recommendation is to follow Adobe’s best practice by ignoring data in these dimensions and instead using marketing channel, referrer, referring domain, and the dimensions associated with Ad Analytics for Paid Search.

To be thorough, and because your installation of Adobe Analytics may be configured this way (it might not be possible or prudent to attempt to change that, at least not quickly), it is useful to provide details on the original goals of these Adobe dimensions. That said, please consider the recommended best practice instead if you are in a position to do so.

Detecting paid search visits with Adobe Analytics

Adobe Analytics provides administrators with the ability to define rules to help differentiate paid search from natural search. The rules are set in a report suite’s Admin Console, listed under Report Suites → Edit Settings → General → Paid Search Detection. One automatic rule that Adobe provides is that a visit must have a referrer that is a known search engine.

Adobe thankfully keeps this list updated so admins do not have to concern themselves with it. The remaining paid search detection rule definitions are based on a query string parameter, for example: cid=PS. Companies can set up different query string parameters based on the search engine, but we’ve found it preferable to use a single variable across all engines to keep data clean more simply.

The image below shows how to configure paid search detection, which mirrors Google Analytics standards.

Adobe Analytics Report Suite
A report suite’s paid search detection mirrors Google Analytics standards.

If you’re familiar with Google Analytics, you’re probably used to the concept of utm query parameters to define marketing channels such as paid search. Google Analytics requires that you use utm_medium=cpc as the query parameter to properly bucket paid search visits. Because Adobe can define paid search based on any query parameter, brands that transition from Google to Adobe tracking can keep the same query parameter. The report suite’s paid search detection rule simply needs to be taught to look for utm_medium=cpc.

Differentiating paid search in Adobe Analytics

The simplest of the dimensions focused on search engine data is the paid search dimension. The paid search dimension helps analysts break down search engine behavior as either paid or natural. This high-level breakdown can be used to easily differentiate behavior at a very high granularity.

Adobe Analytics paid search table
A freeform table shows the simplicity of the paid search dimension.

Analyzing paid and natural search engines in Adobe Analytics

All behavioral data from all search engines, regardless of paid search detection, is tied to the search engine dimension. The dimensional values are thankfully friendlier than just domains. Adobe returns the data as text, such as Yahoo! or Google — Denmark.

Adobe Analytics search engine data
Search engine is shown with a paid search segment applied.

The friendlier view of your search engine data can be useful when filtering or segmenting data to find exactly the engines you’re trying to analyze. The image above shows the search engine dimension with a paid search segment.

Do you see anything strange in the image above? Because the data is sorted by visits, which doesn’t have a segment applied to it, the first line item is listed as Unspecified.

Unspecified is listed at the top because it is the result of all the visits that didn’t come from a search engine. If an analyst were to sum all the visits to each of the individual search engines, there would be a significant difference between that sum and the total count of visits to the site; Unspecified acts as the remainder. Adobe adds an Unspecified row by default for almost all dimensions to make it easier to focus on behavior where the dimension was not set (or unspecified) when that metric was captured.

Adobe makes it easy for analysts to remove that dimensional item from view through the table filter feature. The image below illustrates the details to remove Unspecified now.

Adobe Analytics advanced filter
An advanced filter is applied to exclude Unspecified.

Paid search detection rules help analysts by creating two dimensions at the search engine granularity: search engine — natural and search engine — paid. The only difference between these aligns directly with whether the visits met the detection rules.

Analysts can use search engine data to help marketers better attribute their marketing dollars. If one paid search engine is driving a significantly higher amount of traffic but a lower conversion rate, it may make sense to adjust the budget for this search engine. Search engine alone isn’t usually enough to make this recommendation. As you would expect, Adobe also provides similar dimensions focused on the search keyword rather than the engine.

Initiating search keyword analysis in Adobe Analytics

Search keyword allows analysts to dig deeper into their search advertising data to identify what keywords are driving prospects and consumers to visit their site. These keywords can often become some of the most useful dimensional values to an analyst; when else do consumers tell you exactly what they’re looking for?

Unfortunately, there’s a catch. Years ago, in the name of privacy, Google blocked natural keywords from view by all analytics platforms. Other search engines soon followed suit, and now our beloved natural search keywords have been removed from Adobe Analytics (and Google Analytics, Webtrends, Coremetrics, and so on).

The search engines did, however, continue to provide advertisers with access to capture the search keyword if a user clicked through on a paid search ad, but only if that keyword as sent via query parameter on the landing page.

So what does this all mean? All three of these dimensions are mostly useless because they generally list just Keyword Unavailable. You may see some minimal data in them from search engines that have not yet blocked paid search, but you should instead collaborate with your Adobe admin team and advertising team to ensure that paid search keywords are captured in a custom Adobe dimension.