Understanding Search Analytics: Making Better, Data-Driven Decisions

Do you feel like your analytics is working for you? Like really working? If you don’t, you’re not alone. In fact, a survey by Covario revealed only 20 percent felt their current analytics setup gave them insight to help make better business decisions.

The cure? Do a better job of leveraging data. Sounds simple enough, but in today’s multiscreen world, sifting through mounds of data and working through proper attribution can be a full-time job.

The good news is that we have more data now than ever before (or is that the bad news?); today, it’s just a matter of getting a handle on what that means to our business and our marketing programs.

These concepts and more served as the foundation for a discussion on “Understanding Search Analytics” at Covario’s INFLECTIONPoint 2014 event, just last week.

Bird’s Eye View of Device Usage and Tracking

The devices we optimize for and track are changing. And that’s no secret. But as marketers, we always need to be one step ahead. For example, the devices that matter today, that our target audience uses, will increase or decrease in importance over time. So where will you continue to focus your tracking efforts?

Covario’s Steve Beatty shared the results of an eMarketer/Nielsen survey that illustrated the importance marketers place on certain devices today, and which devices they felt would be important a couple years from now:



This data segued nicely into a presentation by Google’s Michael Burke, who spoke about optimizing the experience for a multiscreen world. And it’s not about the device, he said.

Take off your marketer hat for a moment, and put yourself in the shopper’s shoes to get an empathetic view of device usage, and how to make that experience great for that user at that moment (referencing Google’s ZMOT philosophy).

Burke walked us through a scenario that went something like this:

  • A guy is thinking about getting a new shaving kit.
  • He decides to search for it on his mobile phone as he sits in the dentist’s office.
  • He sees something interesting in the results, clicks through and then abandons the process.
  • Later, he decides to continue the research at home on his tablet, and he finds a store nearby that carries the product.
  • A couple days later, he goes to check it out. While there, he does a price check on his mobile phone to see if he can get it cheaper elsewhere (aka “showrooming”). He sees he can get a discount online.
  • Finally, a few days later, he makes the purchase on his laptop at home.

If you can tailor the experience to the needs of the user when they are performing search or engaging on your site on a particular device, you’re beginning to think in line with ZMOT’s philosophy of sending the right message at the right time.

And it’s not just devices that matter. The intent behind the action is worth understanding to become even more relevant to your audience. For example, Burke said, someone searching for “best mobile phone” on a Wednesday night on a laptop has a completely different intent than that same search coming from a mobile phone on a Sunday afternoon.

So when you’re optimizing for devices, optimize for the experience.


And with this, comes the need to attribute not only how devices impact the experience and the bottom line, but also how each action (no matter how small), can lead to something bigger. As Burke pointed out, ecommerce transactions are not the only value-driven action. In fact, he said, there are lots of untracked events that have serious positive consequences.

But there’s still a ways to go when it comes to measuring the value of certain actions. In-store purchases are one conversion point that Burke said Google is working hard towards finding ways to attribute which actions drove that purchase, and, he said, they’ve started experimenting with Web history and geolocation to begin to get an understanding of that.


OK, So How Do You Attribute?

All this talk about attribution sounds great, but not everyone has a handle on how to do it. As Covario’s Anthony Stagg pointed out next, attribution is still an emerging discipline, but a lack of standardization plus challenges with online data access, collection and quality can be an issue.

Especially as we face a “cookieless” world, as illustrated by the following graphic:


There are implications, Stagg said. Consider the following:

  • Every cookie generates its own user profile leading to multiple and often conflicting profiles of user behavior. Cookies don’t work with smartphones and tablets.
  • Google, Microsoft and Facebook have begun development on their own browser-based individual identifiers.
  • These individual identifiers would allow them to collect behavioral data across all content, campaigns, and devices—smartphone, tablet, PC and laptop—and attribute them to a single unique user profile.
  • This increases the threat of Internet giants monopolizing online user behavioral data and selling it to advertisers and agencies.

Today, when companies do practice attribution, the majority of it is pretty straightforward. In a study published by Forrester Research, companies most used channels like paid search and SEO in their attribution calculations:


And even though marketers are working towards more sophisticated attribution modeling, the last click, said Stagg, is still the default for many.

His recommendation? Explore the technology that’s out there now, looking out for the many new features and functionality attribution technology vendors have to offer.

Drowning in Data? Investigate the Numbers

Half the battle is getting the data; the other half is interpreting it. Covario’s Nick Morrelli pointed out the pitfalls of analyzing data next. One such pitfall is treating the data you collect site-wide in the same manner.

For example, it’s common to treat your website users differently based on where they engage on the site, but Morrelli argued it’s time to start analyzing the data for those pages differently as well.

One scenario is the bounce rate. The bounce rate for users in your sales segment should be looked at differently than the bounce rate for your users in your technical support section, he said. In those cases, the bounce has different implications.

If you’re drowning in data or your numbers just feel lackluster, there’s a remedy: dig deeper. One of the best things you can do when analyzing data is to always ask yourself, What is the next step? What other questions can I answer?

And to avoid becoming too comfortable with face-value analytics; Morrelli reminded the audience to always look for evidence to prove or disprove your assumptions.

As Beatty took the stage again for closing remarks, he left us with a quote from the fictional detective character, Sherlock Holmes: “The temptation to form premature theories upon insufficient data is the bane of our profession.”

As marketers, it’s easy to become overwhelmed with data, and do nothing to resolve tracking what matters. But the investment up front in doing so can pay dividends to making better data-driven business decisions through search analytics.

Related reading

SEO strategy website redesign migration
siddharth taparia SAP on embarking on search transformation projects
How to improve SEO using data science
Four tools your business needs for better metric reporting