• by Ryan DeShazer , October 14, 2011

I am an admitted data nerd. I get excited about funnel visualization. I hang on Avinash Kaushik’s every word.

Coming clean about those truths may not make me unique in the search space. There is a certain science to our craft that can only be validated with data’s cold hard facts. I can’t imagine what search would be without data. It certainly wouldn’t claim the share of marketing investment it does today without the proof provided by the underlying data.

My journey with data began the first time I used the excellent, and now extinct, ClickTracks Analytics (may it rest in peace). Anyone who was into analytics during that time knows John Marshall, the company’s quirky front man. I used to watch his “Marshall Law” video clips religiously; those were weekly shorts on Web analytics best practices, where John wore an Old West sheriff’s hat. I can’t make this stuff up.

Since then I’ve studied data analytics from a number of vantage points, but primarily one focused on Web analytics. I’ve read several books (and have many books yet to read) on analytics theory and approaches to data visualization. The thrill for me is in uncovering the hidden storylines that lie in wait of a savvy analyst to discover. I am indeed a data nerd.

My most recent outlet for geeking out has been Google Analytics. It shouldn’t be necessary for me to discuss how popular Google Analytics is, or why it’s important for any digital marketer to become fluent with its interface and feature set. That should be obvious.

But for the advanced user, Google Analytics can wield some deep reporting prowess. I’ve uncovered four gems in my investigation of Google Analytics’ current feature set. Some of these features have been around for a while; others are new with the introduction of Google Analytics V5. All are essential if you want to take your implementation to a new level of understanding.

Nerddom Stardom awaits.

Events as Goals

“Events as goals” is a new feature within Google Analytics V5. Anyone who has ever managed a digital marketing campaign where online video views or PDF asset downloads were important knows how crucial this new feature is. We could previously track those types of interactions, as either virtual pageviews or “Events,” but we couldn’t set them up as conversion goals. Now with that ability, we can cleanly pull action/conversion reports from a single source, roll up all actions into a single metric, investigate paths to conversion through funnel visualization, and even assign dollar values to specific site actions. Awesome, right?

When used as goals, Events should typically be thought of as “micro-conversions,” or instances of user engagement ahead of the actual conversion (an ecommerce sale, for example). Think of them this way in order to see the full range of influential touches on the path to ultimate conversion.

Multiple, Customizable Dashboards

Another new feature in V5, the flexible dashboard engine has been our most used new tool. Previous versions of Analytics had a very limited dashboard capability; users only had a single dashboard that could be created, and were limited in the number of widgets that could be placed and customized there. Now, virtually any metric or key performance indicator (KPI) can be placed on a dashboard that’s specific to any audience. The dashboard widgets even offer four different ways to visualize the data: single metrics, pie charts, line chart timelines, data tables (but where’s the funnel?!).

Dashboards are crucial when reporting out website- and/or campaign-specific KPIs. While these won’t replace program reporting, they do offer a bit more real-time insight into performance. A best practice is to create custom dashboards for various marketing stakeholders, each unique to their respective needs. When those users log in to Analytics, they’ll find a clean view of the data that is most important to them, without any noise.

Advanced Funnel Analysis through Regex

A regular expression (or “regex”) is a powerful way to programmatically combine strings of text, words, or characters to form a single unit of measure. For example, if I wanted to automatically combine known misspellings of my agency’s branded keyword, I could write out something like:

“gyro|giro|jiro|gyrohsr|gyro hsr|gyro-hsr”

Then when I investigate my analytics account for that particular dashboard widget or custom visitor segment, I would find the total roll-up for visits and on-site actions for that group. I don’t have to look up each derivative individually; regex helps get the job down pretty easily.

Now here’s where the power of regex gets really exciting.

Many organizations view its online purchase/conversion process as linear. The visitor comes to page 1, then 2, then 3, then buys. Simple. Maybe too simple.

In our experience, the conversion process is rarely linear. There are near-infinite paths a user may take on his journey to conversion. However, there are often several pages that are on the website for users within a common mental “stage” of purchase consideration. Customer testimonials are a great example. If my visitors view four different testimonials, they’re likely in an active consideration mode for each page viewed.

Using regex, we can combine each of those pageviews into a single step of the conversion process when specifying a path to conversion within Analytics. Then when you analyze the resultant data, you’re armed with a clearer view of audience needs and how they interact with content at all stages of consideration, and where there are leaks in your funnel.

Rank at the Time of the Click Using Custom Variables

I’ve long believed that “rank at the time of the click” would one day eliminate the need for SEO rank reporting. After all, who cares where a website ranks for a given keyword term if it generates zero click-through?

I even challenged search analytics vendors to make this a new, default feature with my column, “The Call for Smarter Search Analytics.” In that piece, I expressed the impact this data would have on the practice of SEO:

“This metric would address the many factors known to influence natural search positioning across the results pages: personalized search, regional biases, new +1 results. SEOs could leverage that intelligence to make smarter re-optimization decisions, based on resultant on-site behavior patterns.”

Google is now passing specific on-page SERP position data through a dynamic variable, “cd=” when the search originates from Google Instant or when users are logged in to Google Accounts. Through the use of Analytics’ “custom variables” feature, this data can now be brought in and examined more closely.

There are many ways to implement this with custom variables, but here’s a great how-to write up from SEOmoz.

The advances in online measurement that Google has made are tremendous. These are four gems that my team and I have come to rely on. Undoubtedly there are many more to be discovered.