This blog post originally appeared as my column in MediaPost's Online Metrics Insider on August 19.

In my last Metrics Insider column, I wrote about the question of how advertising works. The column generated some great comments, and I got some very thoughtful responses via email; some of you even sent me papers on different components of ad effectiveness measurement. This week, I want to revisit the topic, and perhaps amplify a few points.

1. Advertising works. There are reams of anecdotal, observational, and empirical data that demonstrates that advertising does indeed work. In the late ’80s and ’90s, I remember comScore’s CEO, Magid Abraham, then President / COO of Information Resources (IRI), publishing landmark work, based on hundreds of TV ad campaigns, correlating advertising and sales in a series called “How Advertising Works.” Magid found that incremental TV weight was able to generate sales increases for CPG brands 50% of the time. In our space, my company, comScore, does a lot of work quantifying the effectiveness of online advertising.

Advertising works in many different, sometimes mysterious, ways, with many variables affecting its performance (e.g. creative, media schedule, purchase cycle, share of voice, etc.) Well-designed research can measure that performance, and the best way to do so is against the advertiser’s objective. Suppose a campaign generates robust click-through, but no measurable branding impact. Did that campaign work? Certainly that depends on whether the goal was driving click-through or building brand awareness.

2. Online display advertising works. At my company, we have done hundreds of studies demonstrating the ROI of different kinds of online advertising. In one case study recently presented, we found that, among consumers exposed to a campaign, click-through accounted for only 10% of subsequent site visits and 14% of incremental dollar sales volume; view-through — consumers exposed to the campaign but who did not click on the ads — accounted for 90% of eventual site visits and 86% of incremental dollar sales. In other words, gauging the effectiveness of this campaign based solely on clicks would have missed 90% of the sales impact.

3. Online advertising drives offline sales. Sometimes people forget that the Internet is not a self-contained ecosystem. We can’t ignore the extent to which online advertising can drive offline sales, something else that can’t be counted with clicks. Using our panel and our ability to link it to offline databases, we have been able, time and again, to quantify the impact of online advertising on offline sales. Even for search advertising, wherein one might be tempted to believe the majority of effectiveness accrues in-session and via click, we have observed that 83% of the advertising impact on sales is either latent (sales on subsequent user sessions; 20%) or offline (63%.)

4. Search and display work better together. Another thing we’ve found is that when an advertiser runs a search and a display campaign simultaneously, the impact (as measured by lift versus a control group) of exposure to both search and display is greater than the impact of search alone or of display alone; in fact, impact of search and display together exceeds the sum of the effects of search and display impact individually. In other words, there is a synergistic effect; add two and two and you get five. And not surprisingly, much of the incremental sales generated by the combined exposure group occurs offline.

5. Display ads online are at least as valuable as display ads offline. Television still commands a significantly greater share of ad dollars than the Internet, at higher CPMs. In my last column, I noted that online ads that can be empirically tied to conversion tend to have greater perceived value than ads whose primary impact is measured by awareness, recall and other brand-building metrics. So let me make this point: the impression generated by one consumer watching a given spot presented within long form online video, in full screen mode, is at least as valuable as an impression delivering the same spot to the same consumer on traditional TV. (I would argue that the online impression is probably more valuable because it is likely to also reach the kind of younger, more tech-savvy and harder-to-reach consumer that is increasingly difficult for traditional TV to deliver.)

As we develop new ways to dedicate on-screen real estate to delivering captivating, engaging ads, whether via banner, rich media or emerging formats, I fully expect online display advertising to become an increasingly important component of the media mix. We don’t require a click-through from a magazine ad or a TV ad or a newspaper ad or a radio ad, and all these impressions are valued by advertisers. Impressions online should have at least the same value, wholly independent of the direct linkage to a click. The opportunity to generate that action online is a profound value-add, but let’s make sure that we properly value the ad before we overlay the value-add. Advertising can have immediate effect, but it can also have quantifiable mid-term effect, and profoundly valuable long-term branding effect. And that is as true for online advertising as for any other medium.

This blog post originally appeared as my column in MediaPost's Online Metrics Insider on June 17.

It's funny how things stay with you.

In the fall of 1980, as a second-half senior at NYU, I was taking a class called Advertising and Media Planning. I'd already begun working part-time at Arbitron, so I thought I knew a thing or two. On the midterm, one of the questions was: "Define Gross Rating Points." I confidently answered: "reach times frequency." But the professor was looking for the textbook answer — "the sum of the ratings achieved by a media schedule" — and my answer was marked wrong. I went to his office and argued, but nothing I could tell him would convince him that, indeed, GRPs are the product of reach and frequency.

I was a cocky kid with big dreams, dreams of one day becoming Chief Research Officer at an Internet Metrics company, so you can imagine how this episode scarred me. If that professor is reading this, then let me assert with all the authority and gravitas of my position at comScore, and as a Media Post columnist: GRPs are the product of Reach and Frequency.

Last week, David Smith argued for the efficacy of GRPs as an Internet advertising metric, and a lively debate ensued in the comment section. The case for online GRPs goes something like this: GRPs put online advertising on equal footing with traditional advertising, thus supporting the migration to, and integration of, online advertising as part of the media mix for more categories, brands and advertisers.

The argument against: GRPs are important to traditional media because those poor slobs have nothing else to measure, but here in the digital age, we can measure clicks per you-name-it, rendering GRPs hopelessly archaic. Besides, why settle for equal footing when our metrics make us better than equal?

Perhaps because of my traumatic experience in college, I find myself in the former camp.

Of course, framing the question as either/or is a false premise.

GRPs can and do co-exist with other, more interactive metrics. In fact, I would even suggest that GRPs are a valuable metric for CPA campaigns — because if I know that a certain creative execution is generating actions against a specific target population, I'd like to know: (1) What percent of that target was exposed to the campaign? (maybe I will get more activities by putting it in front of more of that target); and (2) What is the optimal number of exposures to my campaign to elicit that action? (Online advertisers care about this, hence frequency capping.) Can I better manage my CPA campaign if I track these metrics? If yes, well, those metrics are reach and frequency, and if I multiply them together, despite what my NYU professor thought, I'll get GRPs. (I should add that even a behavioral target is, indeed, a target.)

There is one other argument against online GRPs that I'd like to dispel: that somehow the Internet doesn't lend itself to GRPs, at least not as readily as TV or print. I know this to be false, because at comScore, we offer a reach/frequency tool as part of our client interface, and about 100 interactive agencies regularly use this tool to plan campaigns for their clients — running, combined, over 20,000 campaigns a month.

Very specifically, this tool allows users to build campaign schedules by allocating numbers of impressions against a target demographic on different sites, and then showing the aggregate, unduplicated reach (and average frequency) of those impressions across sites.

If an advertiser wants to use online advertising to get a message in front of women 21-49 with kids, and to reach 70% of them at least three times, the tools exist to enable that. I can't think of a single good reason to discourage the practice. Indeed, the entire digital space should be encouraging the practice, because that is the way an awful lot of ad dollars are spent today. The state of the economy is such that competition across media for ad dollars will be increasingly fierce. Making Internet advertising more user-friendly for the people looking to spend heavily to get a message in front of a target audience — that's a no-brainer. The GRP metric and its component parts does nothing but help.

David Smith noted: "We do not generally have easily accessible reach and frequency data for the Web." I think I know what he means — campaigns are ultimately priced based on served impressions, and while we may be able to create pre-buy campaign metrics as readily as in other media, post-buy evaluation is still held to rely on reporting by third party ad servers; thus, it's prone to all the limitations implied by cookie-based tracking.

I want to point out that many advertisers are already using tools, such as comScore's Ad Metrix, which integrates audience measurement data with ad occurrence data to provide post hoc campaign-level reach and frequency.

I don't think there are obstacles preventing advertisers and agencies from planning online media campaigns based on reach, frequency and GRPs, should they be so inclined. Nor should publishers be discouraged from selling inventory that way, as long as there is a market voracious for cost-efficient, targeted brand advertising.

This blog post originally appeared as my column in MediaPost's Online Metrics Insider on February 5.

Last week I was meeting with a client (Hi, Marlene!) when we started talking about the demographic composition of an entity's audience for Unique Visitors (UVs), as opposed to Page Views (PVs). As you can imagine, when you're chief research officer at comScore, you go to a lot of meetings where clients want to talk about... how can I put this delicately... let's just say, when they love their numbers, they seldom call.

Sometimes when talking with clients, I hear the concern that their site targets a specific demographic niche, and yet that niche comprises a disappointingly small portion of their UVs. Invariably I'll hear something like this: "But 85% of our registered users are left-handed Irish backgammon players aged 45-54!" The implication being, the profile of their UVs skews somewhat less targeted.

But here's the thing. The UV metric is democratic to a fault. Every visitor -- the accidental tourist who hits the site once for thirty seconds ever, and the core user who spends an hour a day there -- counts once and only once in the UV. There are a lot of things that can affect a site's UV demographic composition; one of them is search. Suppose that essence.com, which targets African-Americans, runs an article about Tiger Woods. It is possible that many golf fans who do not happen to be African-Americans will end up at essence.com that month because they searched for "Tiger Woods." This search-generated traffic will contribute to the UV metric, even as it dilutes the demographic target.

A better gauge of a Web entity's user profile would be to look at the composition of PVs, because heavy users will drive PVs and tend to counteract the diluting impact on the core target that a UV metric can have. In December 2007, for example, Media Metrix reported that 62% of Unique Visitors to aarp.org were age 50+ (eligibility for AARP kicking in at age 50); but 77% of their Page Views were accounted for by persons 50+, and 79% of their total minutes.

But there is another important reason to think in terms of pages when assessing a web entity's audience make-up. Ads are distributed across pages, not UVs. The more pages one consumes, the more ads one is exposed to, and the more likely that consumer is to see your ad. If an advertiser runs a campaign on a site, the audience profile of the exposures to that campaign will tend to mirror the profile of the Page Views, not the Unique Visitors. In the AARP example above, then, let's restate thusly: 62% of the aarp.com unique audience is comprised of persons 50+, but these persons see almost 80% of the ads.

With a behavioral targeting as opposed to demographic targeting construct, the difference can become even more pronounced. A specific automotive shopping site (here I've chosen to mask the property) had 4.2 million UVs in December 2007. 12% of these UVs (roughly 500,000) came from among the 5.3 million online users who were among the 20% heaviest visitors to automotive manufacturer websites in that month (and who are thus logically highly likely to be in-market for a new vehicle.) But 24% of their PVs, and 25% of their total minutes, came from these "auto intenders." One in four ads at this site will be seen by someone who is in the heaviest 20% of users of automotive manufacturer sites in the same month. When we expand our behavioral target to include the heavy and moderate users of automotive manufacturer sites, then about half the ads on this site will be seen by automotive intenders.

Two final points. One, when I talk about looking at the composition by Page Views here, I really mean, look at composition based on a measure of total consumption as opposed to the total UV (or "cume") audience. If you are concerned about the impact of AJAX on the efficacy of the PV as a metric, the same principle applies with respect to minutes as with pages. In radio, for example, the 36% of a station's cume who comprise the station's core audience consume 72% of that station's quarter-hours of listening.

And two: How about those Giants?

This blog post originally appeared as my column in MediaPost's Online Metrics Insider on January 7.

Over the holidays, everyone (by which I mean, my wife's family) was asking me, "What exciting new developments can we expect to see in online metrics in 2008?" From my perspective, I do see something important bubbling up on the Internet, something that is going to affect online metrics because existing metrics are insufficient to measure it.

I'm talking about video.

I've been in meetings over the last couple of months with some of the biggest general market agencies in the world — the kind of agencies that make the bulk of their money on TV, that typically leave interactive to that digital shop they own out on the coast. Suddenly the Internet has become front and center for these guys, and the reason is video.

It seems like every morning when I scan the MediaPost headlines, some big new video/Internet convergence deal has just been announced. Half of all U.S. ad dollars are spent getting TV commercials in front of viewers, and that pool is poised to spill over into the Internet. In fact, it is not difficult to envision the day when the Internet has become a TV distribution channel, and maybe even the preeminent one — with the flat-screen, wall-mounted monitor in your living room wholly agnostic as to whether the episode of "Lost" you are watching came from the local ABC affiliate, your DVR, on-demand from your cable head end, or streamed from the Internet.

But let's just focus on 2008.

The Internet is not a medium; rather, it is a bundle of media sharing a common distribution platform. I can read the paper online, watch TV online, listen to online radio. I can visit a Web site, use a widget, search for a plumber, or check my Facebook page. I believe that as each of these things emerges as a platform for advertising, we are going to find that one measurement solution does not fit all, any more than a magazine has a program rating, or a radio station has an average issue audience.

I should also point out that even online video is going to become at least two distinct media; for now, let's call the split "TV on the Internet," and "consumer-generated video." All the debate about how to monetize 2-minute clips of guys putting Mentos into Coke bottles, that falls under consumer-generated video, and I'm not thinking about that right now (and I'm not sure that putting home movies on the Internet turns them into viable advertising vehicles.) I'm focusing on TV on the Internet, and I'm pretty confident that the networks will figure out how best to include advertising in their video content, such that with the right metrics, monetization will follow.

So what kind of metrics are necessary to support a robust marketplace for advertising in TV on the Internet? Obviously, the measurement moves away from page views and toward something more duration-oriented. I think the relevant inputs into measurement will be:

• Start and stop times (the building blocks of duration metrics for all audio and video media).
  
  
• Content identification, at the program level, and at the episode level within programs.
  
  
• Program source: to whom should a stream be credited, and how should discrete streaming sources be rolled up into the appropriate ad sales entities? (At comScore, we use something called the Client Focus Dictionary to make these aggregations for urls, pages, and Web entities.).
  
  
• Commercial identification. I think we have two choices here. We can identify the commercials embedded within a stream discretely via some unique identifier; or, we can take second-by-second audience and cross-reference it against a log of occurrence time for ads within the video stream. If a spot airs from the 12-minute mark to 12:30 of a given episode of "Lost," we attribute the audience during that time period to the spot. This latter approach is essentially the way it works in TV, but I'm betting online we align around the first model.

• Who's out there watching? As I've written before, measuring those people out there in front of the screen remains the true challenge for online metrics. But we'll need to do it in order to facilitate online video ad sales markets.
  

I suspect that, at least with respect to advertising and media, video is going to be a major game-changer for the online business in 2008. Is anyone else out there thinking about video? How do you see metrics emerging for the measurement of online video?

Advertising, it turns out, is a lot like physics; they’re both all about time and space. Newspaper, magazine, billboard, and place-based advertisers buy space; TV and radio advertisers buy time. The online business originally aligned around a spatial construct (banner ads defined by area with two dimensions, and existing on a page); but increasingly, online advertisers are migrating toward a temporal construct, especially with the advent of online video.

Historically, Media Math has been pretty simple: “how many,” and “how much?” “How many” is the reach, or the cumulative unduplicated audience (cume), of a media vehicle or ad campaign. Online, our measure of cume is the Unique Visitor. “How much” is tonnage of consumption. In the mid-90s, the online advertising business aligned around the construct of the Page View as the atom for counting Internet consumption (Unique Visitors consume Page Views). In network TV, by contrast, the atom of tonnage is the average minute.

True to the tautology of Media Math, Unique Visitors multiplied by Page Views yields Gross Impressions.

If you followed all that, congratulations! You too can be a media researcher—or just look like one.

Lately, of course, there has been some controversy around the efficacy of the Page View as the atom of Internet consumption—largely triggered by the penetration of AJAX technology, which enables content to refresh on screen without serving a new Page View. In early July, another online metrics company announced that it would stop providing Page View-based rankings, leading to a spate of articles in the business press like this one from Forbes, asking “Is the Page View Dead?”

Meanwhile, everyone even peripherally interested in advertising knows about the industry’s love affair with engagement: two years strong and starting to look serious. With apologies to the ARF, I’m not sure anyone has developed a consensus definition of just what engagement is; we can all agree, though, that whatever it is, we want it. And we’re darned sure we want to measure it.

Engagement is a powerful idea because it represents an attempt to get at the quality of exposure. Not all exposures are engaged exposures, and we want our measurement of ad media to account for the difference. It goes beyond “How many?” and “How much?” to ask, “How good?”

These two developments—the decline in the efficacy of the Page View and the increased demand for engagement—dovetail nicely in Internet metrics, because both argue for a reconsideration of time and space. Specifically, I contend that they argue for a shift in emphasis from Page Views to duration-based audience metrics. Which, in turn, requires some thinking about how consumers, fundamentally, pay attention.

As anyone with a teenager in the house knows, today’s consumers live in an era of multi-tasking. We’re watching Lost on TV with the phone tucked between shoulder and ear while reading email and conducting three chats on IM. Developing a full picture of consumer online behavior now requires that we capture different kinds of media interactions that occur at the same time—and the best way to measure time turns out to be, by measuring time.

Maybe we need to think about online media consumption in two flavors: Time Spent, and Engaged Time Spent. What if we could track the time consumers spend with each web property—whether comprised of pages, audio, video, IM or widget—in a way that allows for capturing multi-tasking behavior? Say I’ve got Facebook and CNN.com open on my screen, along with two IM windows and a ballgame. My engagement to any one of these things ebbs and flows, but I’m spending time with all of them simultaneously. I hop to an IM window and trade messages with a friend, and for two minutes, I’m engaged with the IM client. Then I click onto Facebook to see if my friend likes the same movies as I do, and for three minutes I’m engaged with Facebook. All the while, I’m accruing time with each property (remember: a TV people meter doesn’t stop counting viewing if the panelist answers the phone or picks up a magazine.)

I’m laying all this out because I think this is where Internet audience measurement needs to go—tracking both total Time Spent (maybe we call it “Multi-tasking Time Spent”) and Engaged Time Spent. With these two metrics, advertisers and publishers could start thinking about things like Share of Time Spent; what percent of a website’s Time Spent is Engaged Time Spent, and how can they drive that figure higher? Do the same properties that accrue the highest Total Time Spent also accrue the highest Engaged Time Spent? How do consumers interact with multiple entities when they share screen space?

Some of you might be rolling your eyes right about now; “The last thing we need is MORE metrics!” But I disagree. Consumers use online media in ways that are varied, nuanced and complex; we need to make sure our metrics keep up. And we need to make sure that we keep developing metrics that are driven by consumer experience—that are customer-focused—in addition to the machine-based metrics that are the bailiwick of the Web Analytics side of the equation.

So no, I don’t think the Page View is dead. But I do think that Duration-based metrics will continue to rise in prominence, and that we’ll probably see more of them. Me, I think it’s about time.

What do you think?

The IAB and the MRC are working on a document with the working title, IAB Audience Reach Measurement Guidelines. The arduous work of crafting and wordsmithing is going on under the auspices of the IAB Audience Measurement Work Group, and the leadership of MRC Executive Director George Ivie (he’s the guy putting pen to paper).

This document, once ratified, will help to sort out some of the confusion in the online metrics space over panel-centric versus site-centric “Unique” estimates. The document essentially provides three sets of reach guidelines: one for user-centric measurement (such as is provided by comScore); one for site-centric server data; and one for third-party ad-server networks. The Reach Definition Guidelines will clarify that all Uniques are not created equal. Specifically, the guidelines make clear that “Unique Cookies” (the Uniques counted by site-centric web analytics providers) are not the same thing as “Unique Visitors” (the Uniques counted by comScore.)

I don’t want to go into too much detail, because the document is in draft form and subject to change; because it is not yet official; and because quite frankly it is not my place to “scoop” the IAB. But this passage from the most current draft I’ve seen proves instructive:

“Unique Cookies (unduplicated Cookied Brows