Attribution modeling: the problem with click-based attribution

Yair Halevi (Spock), Chief Architect of SundaySky, wrote a nice piece in today’s iMEDIAconnection on the topic of attribution modeling. This is a subject that really appeals to the geek in me.

Amazingly (although totally believable): “A recent study by Econsultancy and Google Analytics indicates that a majority of marketers are still using last-click attribution to measure performance-oriented campaigns.” The promising news: “…marketers are realizing that it is ineffective at gauging the true influence campaigns have on the consumer’s path to conversion.”

His solution, which makes a lot of sense to me (at least the parts that I understand): “Use control-group analysis together with attribution modeling, not instead of it. Attribution models are still important for analyzing behavioral segments, optimizing the campaign via creative selection and smart targeting, and understanding cross-campaign interactions using control-group testing. However, using control-group testing can take much of the guesswork out of attribution modeling. If you employ both tools, you can use the results of control-group analysis to tune and validate your attribution models.”

BUT, like many articles that I’ve seen on the topic of attribution modeling, it completely ignores the impact of non-digital channels, whether that be traditional media channels like TV or print, the physical retail channel, real world word-of-mouth, etc. I fully realize that this is the Holy Grail of attribution but it needs to be acknowledged at some point inthe conversation. I am not in any way saying we shouldn’t pursue excellence in digital attribution modeling – we absolutely should – but it’s not the best answer to the question most marketers are asking.

Photo: courtesy of http://www.optimine.com. Please contact me before you have me arrested for using this image illegally…many thanks in advance.

Privacy and Industry Self-Regulation

courtesy of GIGAOMI firmly believe that consumers benefit when the ads they are exposed to as they surf the information superhighway are (highly) relevant. And relevance increases in direct proportion to the amount of data we (the industry) are able to collect.

But why are we so afraid that we won’t be able to convince people of this fact? Aren’t we supposed to be experts in changing perceptions…and ultimately behaviors?

If we don’t move to an opt-in (do not track) versus opt-out stance on data collection (a la Microsoft), we will lose any and all credibility to self-regulate. And then the slide down a very slippery slope begins.

Am I being naive?

I’d love to hear your thoughts.

Cross-Platform Video Measurement

Stuart Elliot wrote a piece in today’s NY Times entitled Tracking Viewers From TV to Computer to Smartphone in which he summarized the efforts being made by companies like AT&T, CBS, Discovery Communications, NBCUniversal, News Corporation, Procter & Gamble, Time Warner and Unilever to “measure consumption of video content in cross-platform ways, whether on TV, online or on mobile devices.” The results of two pilot tests will be revealed tomorrow at the Audience Measurement 7.0 conference in New York.

Kudos, this is an admirable initiative and a long time coming.

I’m hoping, however, that there are more compelling insights delivered tomorrow than the highlights cited by Mr. Elliot in today’s article.

“Both pilot tests had a similar major finding: that the growing viewership of video online and on mobile devices is not diminishing the appetite for watching television.” No kidding. This is a fact that has already been confirmed time and time again by multiple sources. TV still captures the lion’s share of media consumption. Got it.

The article went on to say that “Among the findings of the Arbitron test that may be surprising was that of all the people who viewed content on all three screens, the largest demographic group was not the youngest.” Again, no kidding. Younger consumers aren’t watching as much TV as their older counterparts. They’re simply consuming more content on two screens, not three.

There was one point that I fully embrace (color me just slightly cynical): ““all the media companies are trying to monetize their content platforms and to be fully monetized, they need to be measured.” Making more money. Yep. Got it.

But as I said, the effort being made by these companies is encouraging and should be applauded. Maybe the next pilot will try to get at the impact of cross-platform video consumption on sales. That might just be deserving of a standing ovation.

An end in sight for category blindness?

In follow-up to my tweet earlier today in response to an article on AdAge, I wanted to write a bit more about an issue that has driven me crazy for a long time…the institutional belief that the ideal job candidate is one that has spent the vast majority of his or her career working in the same category as the position they are being recruited to fill. The presumptive logic behind this thinking: category experience breeds category expertise. And who doesn’t want to hire a category expert?

Well, me for one. Or anyone else who craves creativity. Dictionary.com defines creativity as “the ability to transcend traditional ideas, rules, patterns, relationships, or the like, and to create meaningful new ideas, forms, methods, interpretations, etc.” So who’s in a better position to be creative, someone who has lived by category conventions their entire career or someone who isn’t beholden to the “tried and true” (emphasis on tried, not true) ways of the past?

I am not in any way saying that knowledge of a category isn’t valuable. It is. But I’d much rather have someone working for me who has broad, horizontal experience across many different categories rather than deep, vertical experience in one. In my view, they are much better suited to deliver smarter, more creative thinking by applying what they’ve learned (and applied either successfully or unsuccessfully) across a variety of very unique marketing challenges.

I’ll fall back on category expertise when I hire a plumber or a mechanic. But not a marketer.

To Pitch or not to Pitch?

There seems to be a little controversy in the old ad community over the new AMC show The Pitch. Should the agencies that decided to participate have done so (see Tracy Wong’s piece on AdAge today for this side of the argument)? For those that decided not to, is revealing trade secrets (the “secret sauce”) nothing but a really lame excuse?

Last week I tweeted a good luck to my former agency who did decide to participate. And I still wish them luck.

But I do have an issue with both sides of the conversation.

For those that opted out to maintain proprietary secrets, please stop. This is a creative business. Creativity happens. And it happens outside (or inside, or beside) any process, structure or trademarked approach. There is no secret sauce. Get over yourself. A great idea is a great idea. Pure and simple. Sure, that great idea has to be grounded in something that matters but last time I checked, no one has a corner on the “brief”.

For those that opted in, what was your true motivation? Very likely it was self-promotion (color me cynical.) Maybe it was the opportunity to pitch a piece of business that you may not otherwise have had the chance to pitch (the chance to make more money)? Agencies are agencies to become famous and make lots of money. And there’s absolutely nothing wrong with that. But ignorance is bliss and as a client (hypothetically speaking) that pays my agency a ton of money to think about my business, do I really want to see them so publicly pitching someone else? Probably not.

So what would I do?

I’d opt out. Not because of some secret sauce. Rather, a belief that you can’t do your best work with a camera in your face…unless you’re paid to do your best work with a camera in your face. And if that’s true, both the prospective client and current clients lose.

And if clients lose, you lose.

Does Mass Marketing Still Have a Place?

I’d like to revisit the broader idea of mass marketing and it’s role, if any, in a future post. But I didn’t want to miss the opportunity to quickly comment on a piece by Dave Morgan (CEO and founder of New York-based Simulmedia) published on Ad Age yesterday entitled “Sorry, the Internet Can’t Fix TV’s Reach Problem.”

Apparently I am one of those that adheres to a “not-so-insignificant school of thought in the media industry that reach on TV doesn’t matter so much anymore.” I must confess that I would much rather find myself in the not-so-significant minority but in this case I can’t.

I certainly don’t disagree with his premise that TV is a highly effective medium (you know, the old sight, sound & motion bit.) There’s plenty of evidence that it works, and works well. Nor do I disagree in theory (I have never seen any evidence pointing one way or another) that the industry tends to talk not walk when it comes to effective targeting on TV.

What I do disagree with is that for TV to be effective, it has to reach everyone. Aren’t we paid to differentiate brands from their competitors by finding unique, relevant and effective ways to engage with consumers? Doesn’t that happen when we understand what makes the consumer of one product different from another and then exploit that difference in where we place our message? That’s not possible in a world where everyone is watching the same programs.

We are a diverse culture (and getting more diverse each and every day) and our media landscape should reflect that diversity. And that means more TV shows reaching fewer people. So where’s the problem? In my view it gives us infinitely more opportunities to be smarter (meaning more effective and efficient) about where we spend our client’s money.

More to come.

Rush Limbaugh, Controversial? No Way!

Was I asleep in my Sleep Number bed when that happened? Or perhaps ordering some flowers online? Granted, he may have pushed his comments a bit far this time but it should come as no surprise to anyone, least of all the advertisers that were more than happy to benefit from his huge audience last week but now under the pressure of progressive activist groups like Credo action have opted to pull their ad dollars from his show. I just find it a bit hypocritical.

There were two especially entertaining quotes published on Ad Age yesterday.

The first, by Maureen Sullivan at AOL: “We have monitored the unfolding events and have determined that Mr. Limbaugh’s comments are not in line with our values. As a result we have made the decision to suspend advertising on The Rush Limbaugh Radio show.” Suspend advertising? Really. So after the hullabaloo dies down (assuming Rush survives this one), you’ll once again bring your money back to support a talk show that’s “not in line with your values?” Just checking.

The second: “‘We have seen some advertisers don’t even know they are advertising on Limbaugh’s show, as they buy the network and don’t know where their spot ends up,’ Mr. Boehlert (senior fellow at Media Matters for America) said.” It’s time these advertisers and/or their agencies started paying a little more attention to where their commercials air. Ignorance is not a good excuse.

We tend to associate with people that we have something in common with, admire and/or respect. And brands should be behaving the same way. Not some of the time. All of the time. Don’t be a fair-weather brand.

The Evolution of Online Targeting

Jonathan Slavin, CRO of CPX Interactive wrote a compelling piece today on iMediaconnection entitled: “The new targeting strategy you should know.” It’s really worth a read.

His basic premise is that the cookie, introduced with the advent of the Netscape Navigator browser in 1997, “has begun to show its age, and while the expectations of online marketers have evolved, the strategies for targeting online audiences largely have not. Now more than ever before, there is a desire for a new targeting paradigm that delivers on the promise of pinpoint focus and scalable reach.”

Rather, he supports the use of IP targeting. Why? “IP audience targeting is an online targeting strategy that brings the strength, accuracy, and flexibility of offline direct-mail campaigns into the online era by mapping IP address ranges to real-world addresses and leverages the publicly available information known about those addresses.” Think income, home value, political affiliation and purchase behaviors for hundreds of products.

He raises a number of very compelling points, not the least if which is the pool of “live” cookies available at any given time. “Forrester Research recently surveyed data service providers and the largest search and ad networks and estimated that “active” cookie coverage at any given time was approximately 34 percent of the online population: ComScore Media Metrix, Inc. estimated 40 percent and JupiterResearch estimated 38 percent. At a high level, this means that from the online population of 200 million people in the U.S., it is likely that no more than 75 to 80 million people are active at any given time.” Insane.

I’ll end by saying that I LOVE the idea of digital kickin’ it old school by adopting the same methods used by the direct mail industry since, well, dinosaurs disappeared from the planet. See kids, you can learn something from your parents.

Flawsome: New Trend Briefing from Trendwatching

I’m a big fan of trendwatching.com and their monthly series of Trend Briefings. Their latest, Flawsome, is no exception (especially as they readily admit that the moniker “is by far our most cringeworthy trend name.” Yep, I agree.)

The overall theme is not a new one but they note that four currents are now converging to make consumers more focused on brand attitude and behavior than ever before:

  1. Consumers’ disillusionment at corporate behavior has (finally) spilled over into outright disgust.
  2. Consumers are more and more aware that personality and profit can be compatible.
  3. Online culture is the culture, and inflexible, bland ‘corporate’ façades jar with consumers who live online where communication is immediate, open and raw.
  4. Last but not least: human nature dictates that people have a hard time genuinely connecting with, being close to, or really trusting other humans who (pretend to) have no weaknesses, flaws, or mistakes – don’t assume brands are any different.

They provide a ton of great examples so check it out. For anyone involved in the management of a brand, it’s a really good idea to remember these points.

Colbert

I am not a regular viewer of the Colbert Report but after watching an episode this past Thursday, I may now be a card-carrying member of the Nation. Colbert makes me laugh in a way that very few others do (which probably says a lot more about me than Colbert.)

But one particular clip – Wheat Thins Sponsortunity – hit particularly close to home as it dealt with all the stuff that I love: brands, sponsorships, product-placement. You really have to watch the clip to fully appreciate it, but Colbert mercilessly lambasted the folks at Nabisco (as he is prone to do, and do quite well.) Why? For taking the Wheat Thins brand a little too seriously.

Now I’m the first to defend the importance of building a brand. A brand can be a critically important and valuable asset. But at the end of the day, it is only a cracker, or a sneaker, or a coffeemaker. It rarely saves lives. If it disappeared tomorrow, the human species would survive.

For those of us who have any role in the management of a brand and its role and reputation in the world, we’d be doing ourselves a service to keep this in mind.

Now I’m off the set my DVR to record every episode of Colbert it can find.