This post was originally published at SearchEngineWatch on July 25, 2011.

Summer travel is a daunting prospect. Everything is packed: the planes, the hotels, the roads, and your cars (especially those with kids). We have all witnessed epic meltdowns by fellow travelers and travel workers alike, the heat and the impatience pushing everyone to their breaking point. Fortunately the Internet has taken some of the hassle out of the process, offering up a myriad of options for research, planning, and booking. Search plays an integral role in this process and tells us quite a bit about both travel marketer and consumer.

Search clicks to Travel sites from the Big 5 search engines are up 10% year over year reaching to 272 million in June 2011. This annual upward search trend tends to start in April and peak in July, so proper planning and execution is critical for marketers this time of year. Summer travelers begin their planning process early and turn to search for information and advice on where they should go and stay, as opposed to lots of holiday season travelers that know where they are going (e.g. to see family, which requires much less preparation and lead time). Marketers must take this into account when they are preparing their campaigns for the season, as there will be many more touch points to reach their consumers during their purchase process.

To further illustrate this point, 50% of the 2011 travel season search growth is driven by clicks to Travel-Information sites. Travel-Information sites are content based sites where prospective travelers can read reviews, learn about different destinations, and engage with their fellow travelers in a variety of fashions (TripAdvisor is a main player in this space). Visits to these sites traditionally happen early in the travel purchase funnel and are great opportunities to work with one of the big search retargeting vendors like Magnetic in order to drive a higher conversion rate. Search and display advertising synergy has been proven to work in tandem to deliver the latent activities that you covet, such as branded searches and conversions. Since the online travel purchase can take some time to develop, the earlier you imprint your targeted brand message into the minds of the travelers on these content-based travel sites (perfect for display advertising), the better your results are likely to be down the road.

One of the big reasons targeted brand messages via search and display are so important early in the travel purchase funnel is because travel searchers are incredibly brand specific later in their purchase process. They may start out their travel planning with a general term like “hotels in Las Vegas” or “flights to London,” but by the time they are ready to book they start typing in branded destinations.

10 of the top 10 search terms driving traffic to travel sites are brands (although “Cheap Tickets” and “Cheap Flights” can have both branded and generic meaning). Online Travel Agencies (OTAs) used to completely dominate travel search share, but have relinquished much of that share to the supplier sites over the years (i.e. Airlines and Hotels). The battle is fierce, and every marketer will do whatever they can to get you to convert during your visit, since they know you are beginning to reach the point of purchase when you visit their sites by way of branded search terms.

Due to how important these branded search clicks can be in converting prospective travelers, most every travel marketer spends a considerable amount of money bidding on their branded terms via paid search. This allows them to own multiple spots on the ever important travel SERP and deliver multiple targeted brand messages to the travelers during a critical juncture in their (search) journey.

Although it may be too late for you to launch a major campaign this late in the season (and with the annual August drop off fast approaching, I don’t recommend it), the numbers point towards a travel search renaissance following a considerable slowdown over the past couple of years. Organic clicks to travel information sites are king these days, but the OTAs and Supplier sites are not far behind. Paid search should be combined with your display advertising strategy and pushed hard with brand messaging by April at the latest.

And most importantly, remember that “cheap” is a bad word… except when it comes to travel. As the 2nd most popular word used by travel searchers every month, it should be in your search toolkit... don’t leave home without it!

comScore is very excited to announce today that it has entered into a definitive agreement to acquire AdXpose, a leader in ad verification, optimization and brand safety. Consistent with comScore’s track record of leadership and innovation, the acquisition of AdXpose, combined with comScore’s existing Campaign Essentials product, will enable the development of game-changing metrics in online advertising measurement. These new metrics offer the opportunity for more meaningful insights that will change the way the digital media industry currently buys and sells advertising.

Online GRPs: Necessary, but Not Sufficient
Recently, there has been a significant amount of discussion about the value of online GRPs, a metric that comScore has been reporting for several years. The analysis we’ve done across thousands of campaigns during that time has led us to an overwhelming conclusion: GRPs, while important and necessary in that they create a bridge to traditional media, are simply not sufficient for effectively managing the dynamics of how advertising works in the digital environment.

Consider the ad delivery process online. Unlike traditional media, as much as 80% of all online ads are delivered through third parties. As a result, the advertiser is often unaware of how, where and to whom these ads are actually delivered, which presents some significant challenges. In some cases, a portion of these ads are delivered but are not seen. In other cases, they appear alongside content that, in the advertiser’s view, would be damaging to the brand. Sometimes, within the complex multi-party hand-offs involved in delivering an ad, unscrupulous players (often unknown to other parties in the transaction) can even game the system and introduce fraud into the delivery.

Even when an ad is legitimately placed next to content that is not objectionable, it is often delivered to people who are outside the intended target, either demographically or geographically. In addition, ads may be delivered on target but at too high or too low a frequency. In the thousands of campaigns we’ve analyzed, it’s not uncommon for 25% of exposed people to see just one impression (which is typically not enough), and 25% or more of exposed people see it 10 or more times (which is typically too high). Collectively, these factors contribute to significant difficulty in achieving the objective of an advertising campaign, which is to deliver ads to the right people in the right context at the right frequency in order for advertising to have the desired impact. The current environment leads to sub-optimal campaign performance and a lot of unnecessary waste.

This reality is not good for anyone in the digital ecosystem — advertisers, publishers, agencies, ad networks, ad exchanges, etc. — and it inhibits the growth and development of the digital medium. What is needed is a way to understand the Validated GRP; in other words, a measure of impressions that were legitimately delivered to a user, in the right context, within the right frequency range, to the desired target.

Digital Media Industry Supports Need for Enhanced Transparency in Ad Measurement
It is not just comScore that has come to this conclusion about the need for Validated GRPs. Recently, the Association of National Advertisers (ANA), American Association of Advertising Agencies (AAAA), and the Internet Advertising Bureau (IAB) collaborated on standards to better report advertising exposure as part of their Making Measurement Make Sense (3MS) initiative. One of the core standards they suggested was to report on viewable impressions in campaign delivery and effectiveness, which would bolster confidence that ads being delivered online were properly seen. The 3MS initiative also aims to reduce the cost of doing business in digital advertising, where the proliferation of media and advertising technology suppliers has made executing a digital ad buy have a greater proportional operational cost than other media. comScore and AdXpose will reduce this complexity by providing a “single tag” solution to measure all of the metrics needed to ensure valid advertising delivery to the right audience. The single tag eliminates operational costs and reduces discrepancies for all parties in the online ecosystem.

In order for digital advertising to fulfill its potential as a strong component of the marketing mix, our industry requires greater transparency in understanding the complex dynamics of ad delivery. Enhanced measurement capabilities are needed to help proactively manage campaigns, and to take corrective actions when inefficiencies occur. comScore’s acquisition of AdXpose sets out to accomplish exactly that by integrating its unique ad validation and optimization tools into our Campaign Essentials product. While these products will initially be sold side-by-side, we are already working to combine them into a fully integrated, end-to-end solution where these metrics can all be delivered, managed and optimized in real-time, within a unified system.

In summary, we are excited about welcoming the AdXpose team and its world-class technology, and we look forward to the new value we can deliver to the industry together. Stay tuned for more developments!

Are online health websites really experiencing a slowdown in visitation? Contrary to recent survey findings reporting a decline (PDF), comScore shows that overall visitation to health websites has never been stronger. Ever since consumers gained unprecedented access to a wealth of information on the Internet, the demand for health and pharmaceutical information has only grown. Now, as connected devices such as mobile phones and tablets enable users to go online with even greater freedom, online health content is poised for continued growth. To better understand the prevailing trends in this sector over the past few years and where it appears to be headed, comScore analyzed several of the key dynamics in the industry today.

Health Site Audience Grows 60 Percent Over Past Three Years
Over the past three years, U.S. Internet users have shown a steadily increasing trend in visitors to sites in the Health category, which range from general health content sites to branded pharmaceutical sites. The number of total unique visitors accessing these sites on a monthly basis has increased from 86.9 million in June 2008 to 139.1 million in June 2011, representing a 60-percent increase.

Interestingly, the rate of growth in visitors to health properties over the past three years outpaces the growth of the total U.S. Internet audience by more than a factor of 4 (60 percent vs. 13 percent), showing the demand for health information continued to increase at a strong pace.

Even more telling is the growth in audience penetration of Health properties over the past few years. Three years ago, less than half of the total U.S. online population visited health sites. Currently, health sites now reach approximately 2 out of every 3 Americans going online monthly, an increase in penetration of nearly 20-percentage points since June 2008.

Key Health Domains and Pharmaceutical Sites Sustain Growth in Health Visitation
While this growth can be attributed in large part to the growing use of sites providing general health information, it’s important to note that it has also been bolstered by visitation to health content domains for major therapeutic areas (i.e. WebMD Depression, Healthline ED, etc.) and pharmaceutical product sites. Over the past three years, targeted health content sites have seen relatively consistent audience levels, even when considering seasonal traffic changes. Yet, a quick look at overall traffic for pharmaceutical sites hints at a slight decline over the past few quarters. Were audiences seeking less pharmaceutical information online over time?

* Content sites include sites in the following therapeutic areas: Acid Reflux/GERD, ADHD, AIDS/HIV, Allergy, Alzheimer's, Asthma, Bipolar, Bipolar Depression, Birth Control, Breast cancer, Cardio, Cholesterol, Cold/Flu, Contraception, COPD, CPG, Depression, Diabetes, Erectile Dysfunction, Fertility, Fibromyalgia, Hepatitis, HPV, Hypertension (HBP), Insomnia, OAB, Oncology, Osteoarthritis, Osteoporosis, and Rheumatoid Arthritis

This apparent decline actually has a fairly reasonable explanation: the launch of Google’s OneBox for Health among search results. In 2010, Google began taking the first unpaid search result listing on pharmaceutical searches to link to the National Institutes of Health (NIH) pages on the medications in question., as illustrated below:

This Google practice redirected a significant amount of traffic that would have once gone to main pharmaceutical sites, causing a decline in search-referred visitation. The chart below illustrates how NIH.gov’s share of visitation from organic search referrals increased from less than 1 percent of clicks to 4.5 percent of clicks after the Google OneBox launched, coinciding with a decline in clicks to branded pharmaceutical sites. Yet it is important to note that any resulting traffic still involved exposure to branded pharmaceuticals – just not on the branded pharmaceutical sites themselves.

This leads us to another key metric to consider when studying the growth in online health visitation – the volume of overall traffic brought to Health properties from web searches. An analysis of these traffic patterns over the past year shows an increase in the volume of search-referred traffic to Health properties, indicating no shortage of demand for health information from search queries. So despite the impact of Google OneBox on visitation to branded pharmaceutical sites, it would be incorrect to assume that the reason for such declines is a lack of consumer demand.

Mobile Access to Health Information Increases Prospects for Growth
Looking to the future of online health content, it is important to understand how the mobile channel is beginning to play an important role. In the three month average period ending June 2011, 13.1 million mobile subscribers in the U.S. reported accessing health information at least once in the previous month, a sizeable increase of 64 percent from a year ago.

A look at how these users accessed their information is also telling, as nearly two-thirds of them used browsers instead of apps to find health information. But while app usage accounts for less overall activity, the growth in the use of health apps outpaces the growth in visitation to health sites via browser (107 percent vs. 84 percent). As smartphone adoption continues to steadily increase, it’s possible that the use of apps will keep on growing at a faster rate over the next few years. What we’re seeing provides a glimpse into a future where connected devices could be increasingly valuable access points for audiences seeking health information.

At the end of the day, these trends we’re seeing from comScore data show the demand for online health information to be far from waning and the prospect for sustained health visitation to be strong. Consumers have never before had as much ability to find health information to inform their health care decisions as they do now, and with the proliferation of connected devices enabling greater access and constant connectivity, it is only likely that the use of online health sources to engage with health information will continue to grow.

In earlier posts, I’ve lamented the sad plight of small retailers. They were hurt far more than their larger competitors by the recession and its aftermath, being unable to match large retailers’ aggressive pricing and discounting tactics. So, it was with relief that I noted some improvement in small retailers’ share of the online marketplace over the most recent three quarters:

However, it’s important to put this positive trend in perspective and realize that small retailers’ 33.6% share in Q2 2011 is still well below their 37.2% share back in Q4 2009. This depressing statistic points to continued weak consumer demand due to high unemployment and concerns about the country’s economic future.

In fact, the National Federation of Independent Business (NFIB) recently reported that its Small Business Optimism Index declined in July for the fifth consecutive month:

Fifth Consecutive Month of Decline
"For the fifth consecutive month, NFIB’s monthly Small-Business Optimism Index fell, dropping 0.9 points in July—a larger decline than in each of the previous three months—and bringing the Index down to a disappointing 89.9. This is below the average Index reading of 90.2 for the last two-year recovery period. Expectations for future real sales growth and improved business conditions were the major contributors to the decline in optimism."

In recent comments, Bill Dunkelberg, the NFIB Chief Economist, sounded particularly bitter:

“Given the current political climate, the protracted debate over how to handle the nation’s debt and spending, and the now this latest development of the debt downgrade, expectations for growth are low and uncertainty is great. At the two year anniversary of the expansion, the Index is only 3.4 points higher than it was in July 2009. And considering the confidence-draining performance of policy makers, there is little hope that Washington will stop hemorrhaging money and put spending back on a sustainable course. Perhaps we might begin referring to the 'Small-Business Pessimism Index' from now on."

Consistent with comScore’s observations in e-commerce, the NFIB reported that the percent of owners citing poor sales as their top problem—the long-time primary complaint of firms—has faded a few points, and reports of sales trends are much better than a few months ago. However, the July survey anticipates slow growth for the remainder of the year, high unemployment rates, inflation rates that are too high and little progress on job creation.

As a result, it appears that we cannot look to small businesses with any degree of confidence that they will help in the near term creation of the jobs necessary to quickly get us out of the unemployment quagmire in which the country now finds itself:

comScore is excited to release a significant enhancement to our Video Metrix product with the introduction of YouTube Partner reporting. This new capability allows comScore to report on the video-viewing audiences cultivated by specific partners within the YouTube universe – an increasingly important component of the online video-viewing landscape.

YouTube is watched by 3 out of every 4 U.S. Internet users each month and accounts for nearly half of all online videos viewed. As the leading means of online video consumption, professional and amateur content providers have increasingly turned to YouTube for widespread dissemination of their videos.

Importantly, this massive video-viewing platform has enabled content providers to amass fairly significant audiences. For example, among professionally-produced content providers, music video channel VEVO attracted an astounding 60 million U.S. video viewers on YouTube in July, while Warner Music had 31 million. The Associated Press’ videos on YouTube reached 6.6 million individuals, which alone would have placed AP among the Top 50 video publishers on the Internet. In addition, Discovery Communications’ videos were viewed by 3.8 million individuals more than10 million times and for more than 700,000 hours.

While these audiences and engagement rates are certainly impressive, professional content providers are accustomed to delivering content at scale. What is perhaps more interesting is how YouTube has facilitated the democratization of content creation and discovery – enabling talented individuals without the resources of large corporations to build sizable, niche audiences. For example, the Smosh duo reaches more than 3 million viewers, more than half of whom are between the ages of 12-24, an average of 8.8 times each month. Revision3’s content creators reach over 7.3 million viewers – with men accounting for 83% of the time spent viewing. Opposite Revision3 stands Alloy, whose audience is predominantly women, who account for over two-thirds of time spent viewing. Also, 75% of how-to video publisher Howcast’s viewing audience has children in the household. The highly-focused audience characteristics unique to these (and many other) YouTube partners are invaluable to advertisers aiming to efficiently reach their target markets.

These YouTube partners and their channels represent important opportunities for advertisers and real monetization streams for content creators. Advertisers can deliver campaigns – whether they are standard in-stream ads, overlays, YouTube’s TrueView ads, or branded content – to desired audiences at scale. The improving economics of online video now offers content creators, (particularly those without the financial backing of large corporations) the opportunity to develop meaningful revenue streams, because these valuable audiences are accessible to, and discoverable by, the agencies, advertisers, and brands that actively use comScore data for campaign buying and planning.

In short, YouTube Partners now have a new strategic ally in comScore, helping them to become regular and pervasive elements of intelligently-designed media plans.

Analytics by Jon Coveney, Jeff Weinstein, and Vinayak Nair from the comScore R&D Team

Maybe we’re complete data geeks, but as we evacuated comScore’s headquarters in Reston, Virginia, a mere 72 miles from the epicenter of the largest earthquake to hit the region in 67 years, the first thing on our minds was wanting to understand the quake’s quantifiable impact on Internet usage.

Throughout the densely-populated and Internet-savvy east coast of the U.S., where the majority of the population has never experienced an earthquake of such magnitude, people flocked to their computers and mobile devices for breaking news of the event. Social media like Facebook and Twitter also played a key role in amplifying the news, which reverberated across the web. While this was happening, comScore’s servers were up and monitoring Internet usage volumes, second-by-second.

Upon re-entering the office following our evacuation, we immediately kicked off a number of scripts to begin aggregating the Internet usage data into usable insights. As we brainstormed, we came up with our newest comScore creation – the “Internet Seismometer” – which is an animated look at U.S. Internet usage by minute between the hours of 1:40pm and 3pm yesterday – for computer vs. mobile-based traffic:

The following timeline highlights what you just witnessed in that video, and the number of immediately-noticeable findings we culled from the data:

• 1:40 – 1:51pm: Business as usual.
• 1:51pm: There is a sudden and precipitous drop in computer traffic in the two regions closest to the epicenter at the time of the quake likely from individuals in office buildings being evacuated.
• 1:51 – 2:00pm: Computer traffic stays low in regions where individuals are most acutely aware of the quake due to proximity but do not have computer access. Meanwhile, other regions have yet to receive word. Mobile traffic spikes during this period, however – increasing in direct relationship to distance from the epicenter: 62% increase for those within 100km, 35% for those within 100 and 500km, and only 7% for those greater than 500km. There is a slight decline in mobile traffic in regions less than close to the quake, prior to this jump.
• 2:00 – 2:10pm: Computer traffic spikes back up as employees in local regions flock back into buildings deemed structurally sound and log on. As expected, computer traffic increases slower in regions further from the quake. Mobile traffic continues to rise across all regions. Peak traffic rates are observed for all devices: PC traffic is up to 22% higher than expected weekday mid afternoon volumes and Mobile phone is up to 28% higher.
• 2:10 – 2:20pm: Mobile traffic declines, especially in regions less than 500km from the epicenter as individuals finish making their emergency calls to loved ones. PC traffic drops slightly but remains significantly high as traffic to social networking sites explodes.
• 2:20 – 2:30pm: Mobile traffic levels off at rates 10% higher than pre-quake volumes. Computer traffic continues a slow decline.
• 2:30 – 3:00pm: Computer and mobile traffic both decrease in all regions, with regions closer to the quake experiencing faster drops as they fall back towards being in-line with other regions. Traffic, on the whole, still remains significantly higher than pre-quake volumes.
• 3:00 – 4:00pm: Computer and mobile traffic hold steady at increased levels, with regions closer to the quake having predictably higher usage.

So, what do these findings tell us about ourselves as members of an Internet-dependent culture? I think there are several key takeaways:

• The Internet provides a means for us to experience a tangible connection to a greater whole. It is a platform whereby the collective effervescence of an event like this can spread like wildfire. And we love that – as demonstrated by how many of us logged on to spur those connections, even in regions far away from the reach of the quake.
• The Internet can mobilize large masses of people unbelievably quickly. In this case, whether it was the thousands of CNN iReports, the tens of thousands of minutes of YouTube footage uploaded, or the hundreds of thousands of tweets, there was one thing in common: we all were talking about the same thing.
• The Internet facilitates lightning-fast mass communication. Tweets from DC reached New York before the tremors did and within less than 20 minutes, Internet traffic had spiked on the West coast to its highest levels. Most people who were using the web because of this event were doing so within twenty minutes. That’s the time it took civilization to get an urgent message to the next town a mere century ago.
• The Internet is for short attention spans. Less than an hour after the quake, Internet volumes were already substantially on their way back down to pre-quake levels. Volumes for today are as would be expected on an average day.
• The Internet infrastructure is not easily taken away from us. The Internet has become a vital connection to both our social network as well as our universal consciousness. When we were evacuated from buildings, we just unconsciously switched to using our mobile devices to connect, uninterrupted.

None of these principles are new. We’ve seen other demonstrations of the Internet’s power recently with events such as the Libyan revolution, Osama Bin Laden’s death, and the Japanese crisis. Thankfully, this latest mass-event left substantially less damage in its wake, but what it did leave was definitive and quantifiable evidence within comScore's data of just how powerful and influential the Internet really is.

This post was originally published at SearchEngineWatch on August 22, 2011.

“Different strokes for different folks”: a simple enough idea to understand and one that resonates with everyone. We are all defined by what we like, what we don’t like, what we buy, and what we don’t buy. Although we are all amazing and unique individuals (like our parents told us), we do share common attributes. Generalization tends to lead us to choosing between two universal identities, often associated with brand names. In other words, “Are you a Coke person or a Pepsi person?”

Search terms can be parsed in a similar fashion. Every searcher can be defined by the words they use when searching. Search engines and marketers alike know this and do their best to deliver you relevant results based on who they think you are and your intent at that exact moment.

As an extension of the targeting by intent strategy, a sophisticated and growing segment of brands are turning to searcher demographics to conduct detailed analyses of their online audience. There are already lots of opportunities for marketers to customize their messaging, placement, landing pages, and the like, for every consumer segment, but the brand managers out there have been using search terms to actually identify the attributes of a “Coke” vs. a “Pepsi” searcher. Considering the money that is spent on brand advertising, knowing how your branded search audience differs from that of the competition should be a valuable nugget of information.

To illustrate the point, below are some fun and interesting universal brand identities with dichotomous stances. Each example analyzes the demographics of searchers that used the branded terms for the month of July, and are based on head of household. The index baselines are the searcher demographics for the entire US search population. Available measures are Age, Income, Location (home/work), Region of the US, Household Size, and Presence of Children in the Household.

Google+ or Facebook?

The most striking differences between Google+ searchers and Facebook searchers are in Age and Income level. Google+ searchers overwhelmingly skew towards 18-34 year olds. Clearly Google+ is a popular brand with the younger segments, and good knowledge for Google to have as they develop their acquisition strategy and evolve their user base. Since Facebook is a much more mature brand in the social networking space, their search audience falls closely in line with the search population at large.

The income skews are even more distinct, essentially polar opposites of each other. More than 32% of Google+ searchers have a household income of $100K or greater, compared to 23% of Facebook searchers. Google+ is definitely off to a fast start in reaching the most desirable income segments, which may make it more attractive to advertisers.

iPhone or Android?

iPhone vs. Android is definitely one of the most frequently debated topics around my office. Which phone do you have? Why that one and not the other? Our San Francisco office almost exclusively has iPhones, and our New York office is littered with Androids. But when it comes to searching for these devices, there is almost no difference between us.

As you can see below, iPhone searchers and Android searchers mirror each other very closely. Considering that searchers of both brands also conduct almost the exact same number of searches for these phones (about 4 per searcher), I would liken these searchers to independent voters. They have not made up their minds yet and there will be multiple times to speak with them before they convert and make a purchase. Knowing that there isn’t a distinct branded search audience would lead me to believe that these searchers are more interested in features, functions, pricing, and carriers, than they would be in defining themselves by the brand, ala Apple Fanboys.

Red Sox or Yankees?

The Red Sox vs. the Yankees is as heated a rivalry as there is today. Both sides have won, both sides have lost, and their respective fan bases hate each other with a passion. Rarely have I ever heard a Bostonian say anything nice about New York City, nor discussed the great merits of Boston with other New Yorkers. But baseball fans are baseball fans, and will have common demographics regardless of the team they support. That said, we do learn a couple of things about their searchers when reviewing the demographics.

Most interesting is where their searchers are based. Both receive a comparable percentage of searches from their own region (43%), but Red Sox searchers definitely stay close to the East Coast. 75% of their searchers are in the New England, Mid Atlantic, and South Atlantic regions. The Yankees, on the other hand, have a more broadly distributed fan base across the country, even receiving 13% of their searchers from the Pacific region.

With this kind of information, teams, ticket sellers, and sports-themed travel companies can better determine where to focus their geo-targeting and messaging efforts. The Yankees Faithful certainly don’t seem to be settling in New England, and Red Sox Nation doesn’t appear terribly interested in places west of the Mississippi.

Like this Column or love it?

Searcher Demographics can prove to be valuable for your business and your brand. It’s one thing to quantify your branded searchers, it is another to know and understand them. The conversation you’ll be able to have with them will only be enhanced. To paraphrase Curb Your Enthusiasm, you will elevate the discussion from small talk to medium talk.

With so much attention given to social networking goliath Facebook, along with strong secondary players in the market like Twitter and Linkedin (and now Google+), it’s no wonder that scrappy upstart Tumblr often gets overlooked. But recent comScore data suggest that maybe it’s about time to start paying closer attention. Tumblr, a multimedia-focused microblogging platform, has emerged as one of the fastest growing consumer-oriented Internet sites over the past year, with its audience surging from 4.2 million visitors in July 2010 to 13.4 million visitors in July 2011 (up 218%!).

As I’ve mentioned in previous blog posts, with social media sites a critical component of success is realizing the benefits of the network effect. The network effect is predicated on the idea that the more users that are part of the system, the more valuable the system becomes to users, which creates a virtuous cycle that pulls more users into the system and gives existing users more incentive to participate. This concept is an important reason why we often see that once social networks achieve critical mass, the network effect takes hold and adoption tends to accelerate.

If we look at the past few months, we are seeing evidence of the network effect occurring at Tumblr, with the number of visitors accelerating from April through July. In just three months, the number of visitors has jumped 5 million for a gain of 61%.

These statistics are certainly impressive, but some of you may still be wondering what Tumblr is and why you haven’t heard of it. If that’s the case, you are not alone – and there is an explanation. To date, Tumblr’s popularity has been strongest among the teen and college-aged user segments. In fact, 50% of Tumblr’s visitor base is under the age of 25. Teenagers age 12-17 are about twice as likely as the average Internet user to visit Tumblr, while 18-24 year olds are nearly 2.5x as likely.

Of course we often see that younger users tend to be early adopters, particularly with social media technologies. So it’s not surprising that the under 25 crowd is currently driving a lot of the site’s popularity. The question now is whether Tumblr can emerge beyond its core demographics and enter into the internet mainstream. If the past few months are indication, it won’t be long before Tumblr is a household name.