How To Find Gold With Web Analytics


goldrush1.jpgADOTAS EXCLUSIVE — A recent Forrester study found that 68% of online retailers say they will spend more on Web analytics this year. Translation: we know Web analytics is important but we’re not taking advantage of it. Unlike social media where marketers are still waiting for concrete metrics, interactive advertisers have hard proof that utilization of Web analytics rolls in the revenue.

Online advertisers are leaving money on the table by under-utilizing or overlooking the impact of Web analytics. Opportunities for incremental gains are not isolated and tapped. Those who do recognize the potential often don’t know what a complete and proper analytics program should look like. What data needs to be gathered? How do you turn that data into actionable intelligence? How do you find the gold?

First things first

While it sounds simple, collecting the right data is where many analytics programs begin to run off the tracks. A few things come up time and time again as inhibitors to the proper collection of data. For instance, an analytics system should be set up to account for all campaigns — visitors generated from an e-mail drop need to be tagged so they don’t look like a direct-to-site visitor. You also want to make sure you track at the most granular level. I often see Google Analytics programs setup to track at the transaction level, but not the individual product level. Assuming your system does not limit you from doing this, granularity is key to finding wins through analytics.

To ensure you are collecting the right data, make sure you set up the right goals. Tracked goals can be prioritized and assigned different values. You’ll want to closely follow the path to a sale or lead conversion, but a newsletter sign-up or white paper download is also a meaningful engagement. By setting up proper goals, you can allocate resources accordingly to focus first on improving those things that will provide the highest return.

Panning for gold

Once a system is in place gathering the needed data, mine that data for red flags and optimization opportunities. There are obvious things to look for like pages with high bounce rates or conversion rate fluctuations, but it is how you dig into those issues that decide what you get from them.

Let’s take the high bounce rate as an example. After a proper setup of Google Analytics, an advertiser finds that a key page within their site has a bounce rate much higher than the site-wide average. The page seems to appear just fine. What could the problem be? The short answer is many things, and in that short answer lays the science of analytics.

Maybe you’re looking at the page in IE and it has a Firefox compatibility issue. A simple browser by browser comparison will point to this. Maybe a look at campaign data reveals a spike in traffic that leads you to uncover that dubious traffic was directed to this page through a paid search campaign. You can then take appropriate steps to modify that campaign and possibly even get a refund. Maybe you discover a messaging disconnect between the campaign driving the traffic and the page itself. In this case, testing ad creatives might better filter your audience leading to more targeted visitors and improved “stickiness” of the page.


The goal of a successful analytics program should be to identify opportunities to improve results. Typically, online marketers are looking for ways to up the engagement level, or user experience, on their site leading to increased conversions and more revenue. Once you have an analytics program that is gathering the right data and helping you identify red flags, you have what you need to start making positive changes.

There is no magic formula for how you move through the gathered intelligence, but it is important that time and resources are allocated to the effort. I like to go with a “highest and best use” philosophy. Where are my resources best allocated to gain maximum return? Typically, you begin investigating one thing that leads to another and another.

A recent client that began a basic analytics program with my agency charged us with isolating opportunities each month to make improvements that would provide lift. After an audit of the setup and necessary changes, we started by looking at a site overlay that told us where people clicked and what the common paths through the site were. Following these various paths lead us to a set of recommendations that ranged from minor messaging changes on certain pages to a complete overhaul of other pages. We were equipped with the data to show that client our recommendations were based on something more than creative differences.

We in the interactive world owe it to our clients to equip them with the tools they need to compete and win. I talk with large companies every day that still endeavor to redesign a site without a true understanding of what is wrong with the current site. Through a small bit of education and an investment in analytics, we can empower them to truly move the needle. Not to mention it is much easier to sell an investment in incremental gains. All we need to do is help our clients find the gold!


  1. Until the analytics includes, parses and attributes offline messaging and communications as well as offline purchase behavior, it will still be fool’s gold.

  2. Excellent article Marc. I couldn’t agree with you more. People /are/ leaving money on the table by overlooking the use of web analytics.

    Simply stated, most don’t even know what it can do for you. Their perception is that the data is too technical and requires a web analytics background. Not true anymore! Take a look around, there are intuitive services out there that present their data in such a way that the small to medium sized business owner can understand and turn into actionable information.

    Google’s “free” service has brought awareness to the process, but I really believe you get what you pay for. There are other solutions out there with a scalable pricing structure that even the little guy can afford.

    Like you said, the key to success is granularity. You’ve got to see the whole picture. The drill down to what each individual users has done is essential.

    Thanks for pointing out how to find gold with web analytics!

    –Tina Bean
    VisiStat, Inc.


Please enter your comment!
Please enter your name here