Analytics Expert

This is the final piece (I promise) in a three-part blog series on web analytics that matters, which — as the previous posts suggested — is a mixture of art (asking the right business questions) and science (having answers that help improve business performance).

The previous post examined how to use custom reporting and tag attributes (meta-data) to uncover hidden relationships between disparate elements in a way that let you capitalize on those relationships. This post will look at two other examples of art in web analytics:

• You demonstrate ROI on things that you didn’t even think you could measure, such as the impact of social media on influencing business-impacting metrics
• You uncover optimization opportunities by benchmarking performance to peers and competitors—again, using metrics that your leadership team cares about

Social Media Analytics: Demonstrating ROI on Impressions

Despite the buzz and excitement around social media, many businesses are not fully investing in social media technologies. This is mostly due to businesses’ inability to adequately measure and demonstrate return on investment. Since social media assets are often deployed on third-party web pages, marketers are unable to deploy web analytics tags and use analytics data to draw straight lines between impressions, or un-clicked social media exposures, to subsequent behaviors and conversions on the marketers’ websites.

Lacking consistent data collection, marketers are unable to tell how hard social media is actually working for them.

To address this challenge, Coremetrics developed a light direct image request tag, called the Impression tag. Coremetrics clients ask publishers to serve a customized version of the Impression tag alongside media assets. The tag works the same way for any type of media — not just social.

In fact, a few Coremetrics clients have deployed the tag together with their display advertising campaigns or embedded within emails, so that they can measure, respectively, view-through and open-rates’ effect on subsequent conversions.

The Impression tag passes data about the exposure to Coremetrics. When a user subsequently arrives at a Coremetrics client’s website, the analytics code will link the impression data with the website behavior data. The client then uses Coremetrics Impression Attribution to attribute website behavior or conversion to media exposures.

While Impression Attribution does not crawl social networking sites, it does tie view-through data about assets such as syndicated videos, widgets and display ads to key business metrics, such as page views, orders and conversion events.

As you can see from the screenshot below, Impression Attribution provides the freedom to ask hard-hitting questions that go well beyond impressions and clicks — questions such as:

• Do syndicated videos drive more conversion events than widgets?
• Are my ads on cnn.com more effective than my ads on nytimes.com — both clicked and un-clicked ads?
• Which version of my ads best performs for my business? (You will answer this question by passing and reporting on impression attributes, such as test version, color, size, location, etc.)

(click for a larger image)

I hope that the Impression Attribution discussion gives you a taste of how far you can go beyond issuing standard reports that trend key performance indicators. With its advanced segmentation and availability of rich metric set, the tool is incredibly flexible. You can truly let your creativity run wild!

Performance Benchmarking

As an aspiring artist, you probably know already that looking at individual web analytics figures, such as sales, unique visitors, and page views, provides only little insight. Knowledge and the ability to act on knowledge become a reality when we put data in context.

In Ronkin’s Pyramid of Context (yes, there is now such as thing), figures are the lowliest of creatures. Web analytics that matters increases reporting value as it moves up the pyramid away from figures, by applying richer and richer context to the reported data.

Moving up the pyramid isn’t easy — the richer the desired context, the more difficult (i.e. expensive) it is to obtain the data. The rewards for moving up, however, increase tremendously, as you get a more vivid and granular picture of the realities of your business, and the actions you should take on the data.

I’ll discuss the different elements of each level of the pyramid in a future post; for now, let’s focus on the apex, which provides the richest level of context — competitive benchmarking.

Competitive benchmarking allows you to compare your business’s key performance indicators against aggregated metrics of peers and competitors.  In a way, competitive benchmarking encapsulates all of the lower levels of the pyramid: You can group figures and ratios and compare them across different periods, where the competitive data act as goals for your business to exceed. Here’s an example:

(Click for a larger image)

Competitive benchmarking allows you to zero in on specific areas of strengths, as well as opportunities for optimization. In a way, it acts like a great leveler, normalizing the data against the impact of seasonality or any macro-level influences, by comparing your data against very similar players who experience the same business conditions.

Coremetrics clients access competitive benchmark data for free, using Coremetrics Benchmark. The solution acts as a springboard, directing clients’ attention to areas they should explore further using custom reporting.

Putting it all together

In this three-part blog series, we’ve been looking at the science and art of web analytics that matters. This was so that you can better understand how to transform your analytics from a finger-on-the-pulse type of practice to a true marketing optimization machine that outputs tangible and significant return on investment. We looked at analytics solution issues that can degrade the quality of data, as well as at the type of reporting that can uncover large optimization opportunities.

I hope that this was valuable to you. I’d love to receive your comments.