This past Thanksgiving weekend was an interesting one for me. I got the opportunity to spend time with my adorable nieces, ages 3 & 5, and since I don’t yet have children of my own, I am constantly learning new things each time we get together. A few of the valuable lessons I learned this weekend include:

1. Children’s books are apparently meant to be read 2 to 3 times in a row, per sitting.
2. Some days kids will just wake up and decide that today is their day to be in a bad mood.
3. Marketing of retail items to children during the holidays – or any time for that matter – works. And perhaps more notably, search activity can serve as a proxy for interest and buzz.

Obviously today’s post is about item #3. Let’s take a look at Zhu Zhu Pets. Have you heard of them? They’re robotic hamsters that scurry around on the ground and make various cooing noises, and they’re apparently on the top of most kids’ “wish list” this holiday season, including my two nieces who barraged me with questions this past week about their likelihood of getting the tiny creatures for Christmas. A recent article from HuffingtonPost.com confirms that Zhu Zhu Pets are in fact the hot toy of the season (“Zhu Zhu Pets Madness” is compared to “Tickle Me Elmo Disease” in the article).

Zhu Zhu pets are actually a relatively new toy – they officially launched nationwide in August 2009 – and the rest, as they say, is history. Parents immediately started blogging about, searching for and trying to get their hands on these furry creatures. If you take a look at the below chart, you’ll see that searches related to Zhu Zhu Pets (including misspellings such as “Zu Zu Pets” and the like) spiked as soon as the toy stormed on to the scene. In August, we see over 129,000 searches for these branded terms, followed by 193,000 in September and up to 234,000 through October. As our November numbers are coming online, the searches and searchers are increasing still.

Parents who searched for and purchased their Zhu Zhu Pets early are likely heading into the holiday shopping season a lot more care free than those who are still trying to hunt down a Zhu Zhu Pet in time for Christmas. The lesson here is not just for moms and dads to start searching and buying early, though. Retailers and manufacturers can also gain a lot from analyzing search activity leading up to the holidays. They can use these insights to try to identify and hone in on the hot toys of the season, allowing them to make informed strategic decisions about production and inventory levels.

The digital world is evolving at a rapid rate, and search marketing is no exception. I was fortunate enough to be on a panel recently at Search Engine Strategies (SES) in San Jose titled: “The Future of Search: Where to Next?”, the purpose of which was to speculate on the roles search marketing might play in the coming years, and how it will evolve along the way. Each panelist was asked to discuss three of their own predictions for the future of search, and I thought it would be valuable to share mine with readers of the comScore blog.

Prediction #1: Searcher Intent Will Not be Judged by Destinations Alone

For the past several years, most of the search industry has held to the belief that Web searchers’ intent can best be determined by the destination of their click. In other words, the theory holds that the best way to determine what a Web user was originally searching for is to examine the site he or she visited after clicking on a link from the search results.

For example, if a person types in “Chicago Bears,” their ‘search intent’ would be determined by the site the searcher chose to visit after considering the various search results suggested. However, the searcher could end up visiting a variety of sites, e.g. the Chicago Bears website to purchase apparel, Ticketmaster.com to buy game tickets or a sports blog with the latest news and updates on the team, etc. As a result, this is not a perfect way to judge the intent, since the exact same search entry can lead to many distinct websites, each fulfilling different objectives. Now, if the searcher conducted a unique search and was led to one unique destination each time, this method has the potential to provide some relevant insight. However, consider the times when the searcher runs only one search query, clicks to visit a site, realizes it isn’t the right one, hits the back button, and clicks on the next link. Using the traditional method of tracking intent, you’d be left with one search query and three possible intentions. Each click would be counted as an intended destination, when in fact the user was only looking to find the final site they visited, (not the first ones they clicked on in error). I have always called this the Homonym Effect of search.

My belief is that, as the search marketplace progresses, search engines and search entities will continue to refine their algorithms in an attempt to produce search results that more accurately reflect searcher intent. Achieving 100% relevance, however, won’t be easy. At comScore, we have also been working to improve our understanding of searcher intent in ways that consider:

• the destination URL (this is important, but not sufficient on its own)
• all of the sites that a searcher visited long before and long after the actual search being analyzed
• all of the types of individuals that use the defined terms

Prediction #2: Search Will be Used More Widely as a Branding Tool

In the not so distant past, online ad dollars made up a very small share of total ad spend across all media. Content with traditional methods of branding advertising, many marketers were not inspired by the idea of venturing into the digital world. According to Lehman Brothers and Think Equity partners, total U.S. measured media spend in 2008 was $186 billion, of which $118 billion was spent on branding. As a subset of these numbers, U.S. online advertising spend was $26 billion, of which $6 billion was spent on branding advertising. As you can see below, even as recently as 2008, only 5% of all media branding ad dollars were spent online:

However, 2009 is showing us that the search market is one of the most vibrant segments of the online channel. As it continues to grow in volume, and as more and more people use search throughout their day, we are beginning to see investments in search advertising by traditional branding advertisers. Companies that had been basking in the comfort of traditional offline media are now beginning to transfer their branding dollars into the online channel.

A great example of this is Kraftfoods.com. Used almost exclusively as a branding vehicle, Kraftfoods.com saw more than 10 million search clicks in the U.S. during the first half of 2009, 33.4% of which were paid clicks (see below). Visitors to Kraftfoods.com can learn about the latest recipes, promotions, and healthy living suggestions before making a purchase of a Kraft brand at the supermarket.

Prediction #3: Search Will Become Increasingly Important Earlier in the Buying Cycle

Traditionally considered one of the final steps in the consumer purchase process, search is moving up the funnel. We are now seeing search activity taking place much earlier in the consumer buying cycle, even in the beginning stages. The number of total monthly searches conducted per unique searcher grew 25 percent in one year (data from June 2008 to June 2009) bringing the total number of searches to 202 million in June 2009. Whether it is an online purchase or an eventual offline purchase, searchers are becoming increasingly dependent on social media and product reviews to assist them at the onset of shopping. The searches conducted to access these tools are also becoming more specific in describing their needs, with the average search phrase now containing more than 3 words per search query.

This shuffle of the consumer buying cycle means that search marketers must be aware of the need to reach consumers earlier in the process. While it used to be typical for consumers to use search after having established interest in a type of product, buyers are now also using search as they establish interest – one of the first stages of the cycle. Given the boundless amount of information that search allows consumers to access, it no longer makes sense to only deliver search ads to consumers when they are ready to make a purchase. It is much more likely that a purchase decision will be made before arriving at a vendor website. Marketers should adopt new strategies to accommodate this change, and opt to market up the funnel to attract customers.

So, there you have it. Those are my three predictions for the future of the search market. Search has already proved to be a very successful tool for marketers and consumers, and it is still growing at an incredible rate each year. As search technology continues to improve, we can expect that search will become integral for many more web functions. Search allows marketers to reach consumers in ways that are not as easy through other channels. The future of search should see an ability to more accurately decipher searcher intent, followed by a push of marketers’ branding efforts online as a result of the use of search by consumers earlier during their consideration and buying processes.

Well, it looks like I came up empty on my prediction about the winner of American Idol last week, so I’ve done some follow up investigation to see where I might have erred. A few qualitative points have come to light that may help with future search-related prognostication:

• The Buzz Factor: Search totals can be a good predictor of interest, but analysis of the terms themselves and the buzz associated with each contestant are also important. Adam Lambert is a polarizing media figure, and the analysis of the positive and negative type of terms used by the Searchers tell an interesting story about the psyche of the American Idol audience. With the South traditionally leading the voting in American Idol, the socially conservative/small town element of the audience was heavily involved. Mark Walsh of Mediapost commented on this approach taken by the folks over at OneRiot, who accurately predicted the outcome.
• The Gokey Factor: I’m kicking myself after the fact for not making the connection.....with only 1 million votes separating Kris and Adam towards the end, the Gokey voters would inevitably throw their support behind Kris for the Finale matchup. Again, I bring this back to the similar style of Kris and Danny Gokey vs. Adam Lambert, the agreed upon “patently original” performer for this season’s show, and if there is anything that we know about crowd behavior, they often fall in line with that which is closest and most familiar.
• The Democracy Factor: Okay, so I wasn’t factoring in that voting for American Idol is not exactly a democratic process. Each fan has the ability to phone-in as many votes as they want, and totals can potentially be further manipulated by the use of auto-dialers. I won’t go so far as to say the system is rigged, but it can certainly be influenced in a way that might not reflect the actual sentiment of the voting public.

So, lessons learned for the next time that comScore lets me be so bold as to prognosticate on the basis of search activity. As Gian Fulgoni said to me the other day, while smiling broadly, “Forecasting is always difficult, but especially when it concerns the future.” ;)

If Americans’ search behavior is any indication which way they’re leaning towards voting for the next American Idol, it looks like it’ll be a runaway victory for Adam Lambert. Compared to other contestants Kris Allen and Danny Gokey, Lambert generated a whopping 78% of all searches for the three finalists during the week ending May 10, according to comScore Marketer data. (Incidentally, Females Age 35-64, who apparently have a track record of predicting the annual Idol winner, gave Lambert a slightly higher 80% of the vote.)

Okay, so maybe predicting that overwhelming favorite Lambert as the eventual victor isn’t exactly going out on a limb. But the search behavior also seems to have picked up on an interesting trend, which is that it shows Kris Allen narrowing the gap considerably with Danny Gokey from the week ending May 3 to the week ending May 10. Certainly an interesting indicator in hindsight considering that the dark horse leapfrogged Gokey to make the final two.