New Video Metrics

According to October data from ComScore and reported by MarketingCharts, “More than 147 million US internet users watched an average of 92 videos per viewer in October.” That’s incredible! The article goes on to say:

• 77% of the total US internet audience viewed online video.
• The average online video viewer watched 274 minutes of video.

Yes, young people 18 – 34 watch more, but we’re all watching them. Here are the charts for number of videos and number of unique viewers.













Hulu, with its emphasis on “long form” video is coming up fast. When I ask groups how many have watched videos lately, everyone says yes. And a lot of them respond that they are watching video of things they weren’t able to see live—time shifting. A final observational note; if you haven’t signed up for a YouTube video channel just to see how a channel works (and think about how it could work for marketers), sign up for the President-Elect’s transition channel and see how they are using it and experience a channel. Don’t worry—you can unsubscribe later, but the fact is, YouTube only sends me an update once a week. A setting probably—I don’t remember--but the point is that it’s not obnoxious.

Back to the point, which is meaningful metrics to understand and learn to make use of all this video that’s going back and forth in cyberspace.

ComScore measures traditional metrics with panel data. As you see from the charts, the basic metrics are:

• Total unique viewers and number of videos viewed
• Engagement metrics such as duration and videos per viewer
• Key demographic statistics for viewers of online video by site and category

If you want to know more, they have good video demos.

Quantcast, which describes itself as “a new breed of measurement service helping buyers and sellers quantify the characteristics of digital audiences against which they can activate addressable advertising solutions.” Translate that: they measure directly, by getting publishers to sign up with them and allow their site traffic to be monitored.

That’s what they’ve done with video. MTV Networks has over 350 digital properties, with the largest and best known including MTV.com, VH1.com, Nickelodeon.com, and ComedyCentral.com, The addition of all these sites to Quantcast’s network of directly-measured properties will certainly add to their ability to provide metrics in the rapidly-growing video field.

Interested in the differences between ComScore and Quantcast data? There’s been some back-and-forth in the blogosphere. Here’s ComScore’s statement and Quantcast’s reply.

The arguments about how best to measure key metrics has been going on since the dawn of mass media, and it continues into interactive media. New developments in metrics are important to all marketers, and we need to keep track of what’s going on!

Engagement Metrics for Social Networks

There was an article today on CNET about start-up BuddyMedia and its new BuddyBrain metrics product. You can see the concept but there’s no sample data yet. The company’s press release provides data from 10 of their clients’ most popular apps:

• The app-vertisements averaged 140,000 installs during the first month of a campaign • Users spent an average of 2 minutes, 35 seconds interacting with applications
• 85% of users who installed the application(s) returned for multiple interactions
• 56% of users returned 9 or more times during the first month

The press release asserts that this is engagement rates 75 times higher than those achieved by banner ads and 5 times higher than with TV. This gives you a sense of the types of metrics they are collecting. Most of their apps so far seem to be on Facebook, so the level of interaction is understandable. Here’s an article with links to several of them. These apps encourage interactivity, that’s certain.

Is there any support for this kind of comparison between Web 1.0 and Web 2.0 media engagement? Not that I could find. But I did find an interesting post by Jeremiah Owyang on an informal metrics survey that he did. It’s a long list and you should read it for yourself. The top 3 items are:

• Meets a business objective
• Supports Community Goals
• Encourage[s] Member Interaction

A more recent post gives the winners of the Forrester social network marketing assessment. They looked at programs from 16 firms. Only one passed using their “ ‘Social’ criteria (rather than traditional marketing tactics) that meet[s] the needs of the community.”

The one “pass” was given to BMW’s graffiti wall app that I’ve written about before. My sense was that it not only had many engagement features, it was clearly centered on the product; this is a marketing program, after all. The BMW campaign scored a 9 (“pass” was 8); Sony BMG’s Alica Keyes progam on My Space scored a 7. Half the 16 firms studied scored 0 or lower!

Most of us have a long way to go in learning how to effectively communicate with our audiences on the social portals and in other social media contexts. Part of the answer is to measure how well we are doing, and BuddyBrain is one approach to doing that. But there’s no substitute for the attitude adjustment that virtually all marketers need to make in order to function successfully in the Web 2.0 world. We have to quit shouting and start listening. Only then will we begin to develop creative ways of engaging with our audiences.

Ask Users to Help Design Site?

From the mid-nineties, when many of us became active on the web and interested in its marketing uses, the question of successful website user experience has been front and center. It’s hard to do, but a lot of people—think Dr. Jakob Nielsen as well as others—preach website usability. Web marketers make extensive use of metrics to understand what content is most popular on their site.

Both are good things. Designing a site for usability should involve a substantial amount of user research prior to and during the construction of the site. That includes asking people what they expect the content to be on various proposed pages of the site. Note, however, that even the research suggests a pre-conceived idea of site structure. Metrics require a site that is in operation with content available. Baynote provides an interesting example of mining user data to better understand content needs, but it’s still after the fact.

What about asking users for suggestions in the early states—perhaps before you even develop the concept statements and concept pages for your site? When I found the DoubleClick ad in the eMarketer newsletter yesterday, I thought “what an interesting idea. Aren’t there other people doing similar things?”

Not as far as I can see. There was a lot of buzz about the “Chrysler Listens” program earlier in the year, and it’s still on the site. According to the Auburn Hills (Michigan) Globe and Mail they have recruited about 5,000 members to their advisory board and are instituting other “listening” programs. This program appears to be mostly focused on product satisfaction, for obvious reasons. VW has a program called “What the People Want” that focuses on popular culture. To me the whole thing seems a bit lame, but it’s drawing traffic, and I have to admit that I’m probably not in the target audience demographic.

So DoubleClick’s program seems pretty unique. The advertising campaign is based around “Three Questions”—answered by a DoubleClick staffer or an outside expert. The viewer can even volunteer to be the outside expert—interesting touch.
Notice the Help Shape our Site box on the main (landing?) page. When you click through, you get 3 serious questions about what you’d like to find in the new “Nerve Center.” I’m always looking for data about our industry, so I answered the questions and submitted my thoughts. They thanked me politely, but didn’t promise anything specific, which probably makes sense in this situation.

How will they analyze the suggestions they are getting? Some type of content analysis, I imagine. That could be a very interesting foundation for site design—content categories suggested by users.

This will be interesting to watch. It’s also interesting to speculate on the impact of Google’s recent acquisition of DoubleClick on this program. But the key point is that DoubleClick is asking users to help frame their new site, apparently well in advance of actual site development. A really provocative use of the interactivity of the Internet!

Media Fragmentation and Advertising Dollars

I recently found a compelling piece of data on media fragmentation. It comes from a webinar and corresponding slides by Christopher Vollmer of Booz Allen. We all know that media is fragmenting around the globe, but data from Vollmer’s new book Always On suggests there is little opportunity to reach a “mass market”—at all, anywhere--these days. Consumer choice reigns, even in that most “mass” of all media, TV.
As a long-time proponent of targeting, I’m not convinced that’s bad news. I do know, however, that it requires marketers to change where they spend their money and how they spend it. There’s a known mismatch between the amount of time consumers spend on various media and the advertising dollars that are directed to those media. Quoting a Forrester study, eMarketer says, “the corresponding difference between time spent online and Internet ad spending was. . .profound, at nearly 4 to 1.” (eMarketer newsletter, February 22, 2008).

Provocative as that comparison may be, it’s not a good metric, for many reasons. Chief among them is that Internet space is still usually cheaper on a CPM basis, which actually isn’t a good measure either.

In an accompanying article Vollmer quotes Carat Americas CEO David Verklin:

“Lately, marketers have become less interested in the number of eyeballs that see a screen or hands that touch a page and more interested in the behavior of the owners of those hands and eyes, and how the ad message connects with them.”

Vollmer goes on to say that “new outcome-focused metrics will shift the focal point of all advertising measurement from exposure to results.” He lists some of the emerging new metrics as:

•Commercial ratings
•Session quality and engagement
•Total viewing behavior (brand contact both online and offline)
•Opt-in activity
•Consumer participation
•Sales impact.

How many of us are actively using those metrics? Do we do the kind of marketing that leads to measurable opt-in activity and consumer participation that ranges from ratings to comments to content creation? Those are sobering questions for many marketers.

Take it a step further. Social media are beginning to fragment also. Consumers have issues with multiple contact points and profiles all over the web, hence aggregators like Friendster and FriendFeed.

Before we even catch up with what’s happening in the “old media,” the “new media” are experiencing the same phenomenon. Difference is, it’s faster in the new media, and the pace of change doesn’t seem to be slowing. Marketers need to choose their channels carefully based on the behavior of their target audience. Then they need to learn to reach consumers in those channels and measure results. It’s a big challenge, and it’s not getting any easier!
Sphere: Related Content

Are Interactive Promotions the Next Big Thing?

A recent report from Borrell Associates, a firm that specializes in local interactive advertising, predicts an interesting shift in the expenditure of online marketing dollars. They assert that display advertising will lose ground to other types of expenditures, especially to promotions. “We believe it will peak this year at $12.6 billion, and then begin a precipitous decline to less than half that amount over the next four years.” The growth will come in online promotions, including public relations activity.

Click-through rates (CTR) for display ads have been getting a bad rap lately (and yes, there’s also been controversy over CTR for paid search ads, but that’s another story). The eMarketer chart (newsletter, February 28, 2008) from the “heavy clickers” study earlier this year got a lot of attention. According to comScore the study found “no correlation between display ad clicks and brand metrics, and show no connection between measured attitude towards a brand and the number of times an ad for that brand was clicked." Ouch!

When Chris Autry asked in iMediaConnection, “Is Display Advertising Dead?” he pointed to low CTR and permission-based content including what he calls narrow networks for special applications and my favorite, widgets. He describes advertising as an application and recommends that, “Consumers will not be bombarded with relatively useless adverts but will instead interact and be able to use a range of applications to perform relevant tasks.”

It is also possible that the movement of traditional mass media advertising dollars onto the web has propped up online display advertising. Will these advertisers accept for long the miserable click-through rates that are the norm today?

What is the growth in online promotions going to represent? According to Borrell contests, coupons and promotional gift certificates are working well at the local level. We’ve seen a lot of evidence of the popularity of video contests in national media, from the SuperBowl to Heinz ketchup.

Are contests reaching the saturation point, though? Ok, I’ll admit I’m not an American Idol fan either, but it seems to me there is a lot of room for creative promotional strategies. If promotion is going to in large part replace display advertising, we need some creative ideas for interacting with our target audience, whether it’s local or national. Any good ideas?
Sphere: Related Content

Who Are the Online Leaders?

Most of us would sadly agree that we’re not one of them. What may surprise you is who some of the leaders are. Here’s some of the headline news that has caught my eye lately:
•According to AdAge, GM plans to have half its $3 billion ad budget into digital and one-to-one marketing within the next three years. GM has been active in the digital space for years with its interactive website, blogs, new wiki and more. It intends to accelerate the move from offline to online media.
•Unilever headed AdAge’s digital a-list for the year. Their campaigns included the various Dove “Real Beauty” programs and a series of webisodes for Suave. Degree deodorant sponsored webisodes for of the popular 24 television program; check out the website and click on the Absolute Protection tab. Unilever marketers are quick to point out that none of these campaigns are purely digital; they are masters of integrated communications using a variety of media.
•Some of the others on the a-list are the usual suspects among agencies and brands like Apple’s iPhone, Google, and ESPN. Others might come as more of a surprise: The NY Times online division and, if you’re not familiar with it, J&J’s baby center.
•Other online leaders like Toyota, American Express and Procter and Gamble are leading the way in the search for metrics that meet the needs of marketers in a digital world.

The corporate names I’m dropping here are not small, innovative start-ups. They are corporate giants and long-time leading advertisers. Clearly, marketers of all kinds are following their leads. Search marketing is a leader in the budget race, both because it works and because it’s easy to measure. An Ad Tech survey says that behavioral and rich media are getting even more budgetary attention. The recent SEMPO survey adds what we pretty much all know; the growth in interactive is coming at the expense of traditional media—print, TV and even direct mail.

In the last few days I’ve also read that another marketer said that digital is beyond experimentation. It is now part of the mainstream media mix, something I've been saying for awhile. There are still questions about how,when, how much and the best executions. But there should be no more questions about WHETHER!
Sphere: Related Content

Customer Acquisition--How the Internet Has Changed It

As I wrote on March 4, the Internet has changed the core marketing communications processes fundamentally and forever. Marketers are still trying to come to grips with that fact and to learn how to leverage and optimize the power of the Internet in integrated marketing communications programs.

We all know what the Hierarchy of Effects looks like. We were weaned on it as marketers. Unfortunately, it’s still the mental model that many of us use. I argue that it is simply not the way marketing works today. It probably was never entirely hierarchical. Today it’s more like a maze with many ways of getting to the end goal—a trusted brand. That makes it hard to specify a process that fits all situations, much less make it one that is hierarchical. Earlier I described it as circular and I think that’s an improvement, but that still doesn’t express the complexity of the decisions marketers face.

In the mass media era we spent time and money to reach our target segment and create brand awareness. In the Internet era the more direct approach is to attract the target’s attention with relevant content. The Internet supports the acquisition process in two significant ways:

•Marketers can target an audience for acquisition with little of the wasted reach of mass media. Targeting by display advertising on carefully-selected web vehicles (sites, blogs, social networks) is similar to mass media and we know that it accomplishes brand development as well as generating action. PPC advertising based on contextual keywords targets to an individual’s current behavior. Behavioral advertising, based on actions already taken by anonymous visitors, offers more precise targeting and is consequently growing in favor with marketers.

•Marketers can incite to action, which usually requires driving people to their website. They must carefully consider the actions they want target customers to take on the website, how they will encourage them to take desired action, and the experience visitors will have while they are there and afterward—in the fulfillment and service process. They must also make decisions about how to measure success and how to capture data from newly-acquired targets.

The set of possible actions represents basic objectives that marketers may choose for their campaigns. They include, not in any order of priority:

•Drive first-time visitors to a retail location to make a purchase.
oThat may be as simple as offering store location information, often with maps and other ancillary information. It can include sales promotions like coupons.

•Encourage an immediate purchase on the site. This can rely on compelling content—from product descriptions to customer reviews—and a well-designed and maintained site that leads visitors through a planned, step-by-step process.
•Provide incentives to make an immediate purchase on the site.
oThe incentives can be part of the advertising—a free shipping offer, for example. They can be presented on a landing page as part of a formal conversion process. They can be presented on the site—an offer to “buy two and get a third for half price” shown, at a minimum, on the home/main product page and on the order page.

•Invite visitors to register by offering relevant content:
oAdditional product information—brochures or demos
oSite functionality—build your own product
oA newsletter or alerts with offers of interest
oCoupon downloads
oParticipate in brand community activities

•Encourage visitors to stay on the site longer
oContent like videos
oActivities like games and contests

•Give visitors a reason to return
oCompelling content, excellent experience, ongoing events

These generic objective types have an element that is familiar to B2B marketers but less so to most consumer packaged goods marketers. They imply a multi-step process, except in the minority of cases in which the first-time visitor makes an immediate purchase. If not, the visitor must be enticed to return. Successful retailers have been good at doing that; producers of mass-marketed products (and some services, insurance sold through agents, for example) have not.

The multitude of possible actions and the fact that not all culminate in an immediate sale pose two additional questions. First, how do we measure success? It is not enough to simply attract visitors to the website. We have to get them to make a purchase. Even though that may take several visits, the process is relatively easy to track on the web. Once it leaves the web for a retailer or a dealer, it becomes much more difficult

Second—and necessary for developing the correct metrics—is what is our working definition of acquisition? Is it merely getting an anonymous visitor to the site? Probably not; that’s the click-through dispute. Is it capturing an email address so you can begin to develop a dialog? If you are marketing a genuinely multi-step product—cars or real estate, for example—registration may be an acceptable definition of acquisition. In those two cases, conversion occurs off the site, so that argues for a more limited definition of acquisition. You may hold out for an initial sale as the only acceptable measure of acquisition. That depends on many things including the product itself and your ability to track through to the sale.

Acquisition is a complex task. The Internet hasn’t really made it simpler. It has, however, made it possible to target--even at the acquisition stage. It has made it possible to measure, not only success in ROI terms, but the path of getting there. Finally, it allows marketers to plan campaigns based on data, both consumer behavior data and programs results data.

The change in approach to acquisition is not an option. Marketers cannot afford to ignore the potential of interactive marketing in their total mix, both for reasons of cost efficiency and because consumers are demanding the relationships.

The changes in acquisition lead to a greater role for conversion. More about that in a forthcoming post.
Sphere: Related Content

Is the Engagement Metric Evolving?

Microsoft’s announcement of its new Engagement Mapping metric at the IAB conference this week created additional buzz around what was already one of the hottest online marketing topics of the moment. And not just this particular moment—it has been ongoing for awhile. Last summer ARF Chief Research Officer Joe Plummer defined it this way, "Engagement is turning on a prospect to a brand idea enhanced by the surrounding context." That’s an interesting concept, but it doesn’t give me any guidance as to how I could measure it.

Several marketers have proposed approaches to measuring brand engagement, especially online. Forrester has a concept that includes four factors--involvement, interaction, intimacy, and influence. That appears to combine attitudinal and behavioral measures, and that would be a strong approach. The public information makes it clear that it requires both online and offline data. That doesn’t make it cheap, but it makes it comprehensive, which is essential.

Brand Keys offers measures of engagement that are category-specific. Their measure relies heavily on customer expectations of brands in the category. They publish a list of highest-scoring brands in various categories each year.

Nielsen//NetRatings has changed its key measure of web traffic from page views to time spent on the site. They tout it as a better measure of engagement, and it certainly beats page views, which have well-known problems. However, that goes back to an old academic argument on the definition of brand loyalty (is engagement a precursor to loyalty--I think so). The argument is that loyalty is more than repeat purchase behavior—that can be just habit. True loyalty—and I suspect true engagement—requires understanding of attitudes as well as behavior.

Earlier this year Kevin Mannion wrote a three-part article (1, 2, 3) for MediaPost’s Metrics Insider Newsletter that gives an excellent summary of efforts to date. He references the work of Eric T. Peterson and Avinash Kaushik, both prolific writer/speakers on metrics issues. Mannion’s analysis of this body of work produces an engagement metric with six components. They are:
Loyalty: how often visitors return to a site over a long period of time.
Recency: how frequently visitors come to a site within a narrow time period.
Duration: how long visitors remain on the site.
Click Depth: the degree to which visitors view site content.
Interactivity: the kinds of actions visitors take with content (downloading content, viewing videos, attending webinars, posting content, etc.).
Subscription: the extent to which visitors register for services or content.
In the third installment he gives an example of how this metric would work. Note that it is all behavioral; to be specific it is all based on online metrics. That means it is based on data that online marketers currently can access.

That brings us full-circle to the Microsoft Engagement Mapping platform. According to Brian McAndrews of Microsoft, “Our Engagement Mapping approach conveys how each ad exposure — whether display, rich media or search, seen multiple times on multiple sites and across many channels — influenced an eventual purchase. We believe it represents a quantum leap for advertisers and publishers who are seeking to maximize their online spends.” I can visualize what such a map would look like and how useful it would be, especially to the multichannel e-retailer.

It doesn’t solve the metrics issue, however. Engagement Mapping is all behavioral—more a measure of impact than engagement in my mind. If engagement is indeed an attitudinal state that is manifested in brand behavior of various types—both on and offline—we still don’t have a metric that truly captures the concept.
Sphere: Related Content