Search Engines by the Numbers

Last week I attended Search Engine Strategies in New York. I took part in the ClickZ track on multichannel metrics with fellow columnist Jason Burby and Web analytics author Eric Peterson. The session covered such areas as building an overall metrics framework, integrating multiple data sources, and building out your key performance indicators (KPIs), themes that crop up regularly Jason’s and my columns.

Having done my bit, I was able to take in the rest of the conference for a couple days. It was difficult to decide which sessions to attend, given the breadth of choice. Since I’m more into analysis than SEO (define), I headed over to sessions in the stats and research track to see what the state of play is in the search market and what the latest thinking is on the way people actually use search engines.

After listening to presentations from companies such as comScore, Nielsen Net//Ratings, and Hitwise, Google’s dominance in the search market was very evident. Various statistics were thrown about, but they all pointed to the clear position Google has as market leader. Not only do more people use Google, but they also use it more often, something we used to call a double-whammy effect in consumer packaged goods (CPG) marketing.

The Hitwise data also demonstrate some demographic differences between users of the various search engines:

• Yahoo users tend to be slightly younger.
• MSN users tend to be slightly older.
• Google users tend to be slightly more affluent.

These differences can lead to different types of sites being searched for on each of the major search engines. MSN sends relatively more of its traffic to business and finance sites than the other two, whereas Google sends more of its traffic to education sites. So there are differences between the users of these different engines. And where there are differences, there are opportunities.

Presentations also highlighted the complexity of user behavior, particularly looking beyond the initial visit. A lot of search campaign response evaluation is based on a direct response model, that is, a click-through resulting in a conversion. A comScore case study on the U.S. travel market showed 80 percent of all transactions occur after the first click-through visit. In other words, only 20 percent of people convert on the initial search visit. They’re likely to return later to complete the transaction, predominantly using a direct entry approach, such as typing in the URL or using a bookmark. The argument is that in some markets, such as travel and other high-consideration products, analysis based purely on a direct response model is underestimating search campaigns’ true effect.

I attended a great session on landing page optimization. This deals with subjects such as A/B and multivariate testing. This is a huge growth area in the U.S. at the moment. Given some of the case study results, it’s easy to see why. Speakers from vendors and their clients pointed to conversion increases of 60 percent based on a single optimized landing page, such as the home page. Though these sorts of gains aren’t sustainable over the longer term, there’s a lot of low-hanging fruit to be plucked when engaging in a systematic program of landing page optimization.

One interesting thing to emerge from the presentations was that in some cases the best landing page may be different depending on where the searcher came from. The best type of copy, graphics, or call to action could differ depending on whether it was a referral from Google or MSN. For me, this tied to the point about different search engines having different user profiles and how that affects their behavior when they reach a site. It’s important to know where searchers are coming from and possibly treat them differently.

Key takeaways from the sessions I attended:

• Search is maturing and Google is dominant, but that doesn’t mean you don’t need to pay attention to the other engines.
• Different engines have different profiles. These result in varying user behaviors on your site.
• It’s important to treat visitors from the different engines as different segments. Understand their behaviour, and develop your campaigns appropriately.
• Different landing pages for traffic from different search engines may be the way to go.
• Landing page optimization and testing are definitely the way to go.
• Campaign analysis based on direct response may underestimate the true value of search marketing activity. A longer-term perspective might be needed. Easy to say, harder to do. One way is to think about where someone came from the first time she visited your site before she converted rather than the last time, which is often the case.

Till next time…

Want more search information? ClickZ SEM Archives contain all our search columns, organized by topic.