Surprising Search Trends: Interview with comScore's Eli Goodman
Search is now the No. 1 activity on the Internet. You need to know how search is evolving and where it's headed.
I interviewed comScore's Search Evangelist, Eli Goodman. Below he offers fascinating insights into how our search habits are shifting and getting more sophisticated.
Using the vast resources of comScore's data sets, Eli teased out revealing patterns and insights that help you and me understand where search is going as an industry and as an activity for the end user.
Search Universe Expands Beyond Search Engines
Larry Chase: When you speak about trends in search usage, you are talking about more than just Google and Yahoo, correct?
Eli Goodman: Yes, that's right. Let's look at some recent research.
Our February 2009 qSearch results showed that in the United States, over 19.1 billion searches took place, with 198 million unique searches, an average of 96.7 searches per person, each month.
These are not just searches in the big search engines but any type of search, such as inside other Web sites, like MySpace, eBay and others.
Google is still the largest search property. YouTube, which overran Yahoo in September 2008, is the No. 2 search property, with 2.6 billion searches in February 2009.
The top search properties are Google, YouTube, Yahoo, and sites owned by companies such as Microsoft, AOL and Ask.com (which includes Thesaurus.com and Dictionary.com). eBay had 513 million searches. After that, you have shopping networks and searches within sites. Facebook dominates the rest of the social networks worldwide for number of searches within the site.
LC: It looks like the average user is doing 3 searches a day.
EG: Well, we segment search activity into heavy, medium and light users. Eighty percent of the searches are done by 20% of the users. Heavy users are doing more than the 97 monthly average searches.
When you think about the concentration of searches, it skews toward heavy users, but medium and light searchers are catching up. Search is becoming more popular, because people are getting more relevant results. We see a 30% to 40% increase in searches year over year.
LC: This is telling me that if the search box on your Web site isn't obvious, it should be.
EG: Yes, but you must also make sure your site's search engine is delivering relevant results to people. Search is driven by which site delivers the most relevant results.
Relevance vs. Privacy
LC: Many times, you have to give up some privacy in order to get more relevant results, especially with cookies that retain browser history. Do you think users will be comfortable with the privacy issues being swapped out for relevance?
EG: We have surveys that address how comfortable consumers are with this [loss of privacy]. What we have found, surprisingly, is that people aren't nearly as touchy as you might think.
People want relevance. There's a real generation gap in people's attitudes about privacy.
Older Internet users aren't as comfortable with privacy loss, but younger users will give it up. If you consider [the information people are willing to post about themselves on their Facebook and MySpace pages], you can see people are much less worried about lack of privacy and are even excited to have something of themselves out there for others to see.
Tying Search Results to Offline Sales
LC: How is search continuing to evolve?
EG: From the consumer perspective, we are seeing convergence of online and offline. As the metrics become more sophisticated, we are able to quantify a lot of latent responses, such as how important different types of ads are, especially with the fragmentation of the consumer decision-making process.
People are doing a lot more heavy research online and then showing up to buy in the store.
I see a lot of organic search growth online and consumers spending a lot more time getting into advice sites and discussion groups.LC: How does that affect the way media buyers work? EG: Imagine someone sits down and does a standard Web search for cars. They start at a general level: "new car," "used car," "used Honda." Over the next few weeks, they will become more specific about what they're looking for.
You don't test-drive 300 cars. You test 3 or 4 and create a short list. This is where the fragmentation of the decision-making process is happening. People are acquiring more information via search to understand what works best.
Three months down the road, by the time they are ready to test-drive cars, they move to vertical search engines like eBay and AutoTrader. This time, they are typing in "2008 used blue Honda under 50,000 miles." That's the analysis a media buyer would need. Which search terms represent which stage car buyers are in?
LC: The longer the search string, the closer people are to pulling the trigger?
EG: Yes. Again, it comes back to the relevancy piece.
For example, before 1998, before paid search, users' results were much less relevant than they are today. Back then, it was easier to game the system to come up higher in search results. As Google introduced PageRank, search results became more relevant, and users became more sophisticated.
Today, we see users becoming more specific by typing in longer search strings. They also are becoming more specialized and more vertical in their searching. This is where you start getting into behavioral targeting and pattern behavior.
Vertical Search Trends
LC: We hear the term "vertical search" thrown around. What does it mean, exactly?
EG: I define vertical search as a search engine that is limited to a specific topic. It could be travel, comparison-shopping engines, etc.
The search market is growing. Do we see search activity moving away from the major engines? No, but there is a lot more search activity within sites that are not traditionally thought of as search engines. The pie is getting bigger, and there are more slices in it.
We track the top 140 search entities. LinkedIn, for example, made an interesting jump in percentages in February 2009. We added it in this year and it is now the 23rd largest search property, based on searches in the US with 19 million searches in February 2009.
Social networks are driving search, such as the searches within Facebook. [Note to reader: We are discussing U.S. search properties here. Worldwide, Baidu is No. 2 behind Google.]
Future of Search and UGC (User-Generated Content)
LC: There's so much user-generated content being uploaded daily; there's no way search engines can keep on top of it. What's the future?
EG: The engines work with the sites where a large amount of user-generated content is being uploaded. They also are learning [user behavior].
Take Google buying YouTube. They are still running in the red on that purchase, but look at the amount of intelligence they can gather on 2.5 billion searches.
What it comes down to is search engines need to categorize information correctly. Any Web site that has UGC has to do this to ensure the data and information uploaded to them [by users] is searchable. The more searchable it is, the more visits the site gets, the more relevant the searches become and the more money the site can make.
The long tail is where the growth is.