Unfortunately in actual practice measurement and accountability often open up more questions than answers for marketers, and create more confusion than clarity. Therein lies the current quandary of online advertising, at least on the brand and display side.

Addressing this quandary is clearly going to be a pressing topic of the next few years, making the heretofore mostly esoteric topic of attribution analysis an increasingly hot-button one.

Most surveys on the topic show that while only a small minority of online marketers use attribution analysis regularly, focus on this class of tools -- whether from ad servers, Web site analytics tools or custom dashboards -- is increasing.

Both Atlas and Dart have been working for some time on "Engagement Mapping," for instance. Nielsen has been talking at various conferences about moving attribution measurement from an impression-based to time-based standard, while numerous firms are working at quantifying everything from buzz to influence, and charting relationships between online ad exposure and offline behaviors.

Another tool in this increasingly populated space is Coremetrics' Impression attribution, a tool for comparatively tracking all impression marketing initiatives including display ads, widgets, micro-sites, syndicated videos across the internet. Though consensus standards about attribution value metrics remains (and is likely to remain for some time) a work in progress, the days when any advertiser, publisher or agency could safely ignore the issue of how to quantify attribution beyond the "last click" are clearly over. Yet at this point most barely know where to start in thinking through the issue.

Key to beginning this process, John Squire, Chief Strategy Officer of Coremetrics, says, is acknowledging and identifying just how much wasted effort and impressions most campaigns actually have.

"Marketers are becoming aware that for all the data they've amassed and all the measurement capability they obviously have they are, they are really flying blind once you get outside the 'last click,'" Squires says.

"Online, more than any medium before it, is a multi-tactical environment," he explains. "There are so many touchpoints and so many potential ways of leveraging each touch point, that for the marketer it's both exhilarating and terrifying. There are so many potentially viable means at your disposal it's possible both to achieve unprecedented efficiency, yes, but also to waste incredible resources."

The fervent hope of publishers and networks, of course, is that better attribution will validate the role of their display inventory. But better validation, Squires cautions, is likely to be a double edged sword. Indeed in the short run, as the ability to actually get closer to a more holistic view of the funnel is actually realized, the role and value of display is likely to come under even more merciless scrutiny than it already has in advertiser budgets.

"Marketers have heard so many claims and pitches over the years," Squires says. "One of the first things that happens is they bear down hardest on identifying what doesn't work, on performing radical surgery on the parts of budgets that lead to wasted impressions. Marketers remain very skeptical of the value of display, so the first impulse of many is just to cut it.

"That, of course, is not the ultimate goal of attribution analysis. The goal is to optimize media spend by locating, first, what does work, and then how exactly it works in relation to every other channel, and how each piece in the mix works, from awareness building and influence to conversion.

"While the first impact of widespread and better campaign optimization is likely to be leaner and meaner budgets for the display side, locating areas where targeted display truly improves overall metrics will ultimately expand budgets.

"A case in point would be a trend we saw about four years ago among some clients who decided because they were scoring so well in rankings on natural search, well, why not cut paid search. We urge them to test that "hunch' for a few weeks before pulling the trigger on paid search. What we found in almost every case was that the tests didn't last past the third day. Shutting off paid search hurt their performance so badly they went right back to it, and we never heard more about cutting out paid search again."

Squires believes that improvements in attribution analysis will be particularly good news for behavioral targeting, especially behaviorally based retargeting. "Very clearly the performance of behavioral targeting is increasingly coming into its own," Squires says. "Marketers have till recently seen BT as a great performer but too 'labor intensive' in that the scale didn't justify the work. But in the past year they are also seeing better scaling. We're seeing, for instance, marketers absolutely restructure their spending mix toward a bigger piece for behavioral display against paid search."

For publishers the challenge will be to learn to better define, package and leverage their user base.

For some publishers, enhanced attribution analysis will expose that publishers have inflated their roles. But for many publishers it will be a godsend in that it truly documents how valuable their audience is.

The stakes, of course could not be higher. For at the heart of the matter of attribution is nothing less than the value perception and proposition of online display ads, the speed of true re-allocation of budgets to online, and the prospects for all the hundreds of targeting platform providers whose business plans are predicated on capturing some of those funds.

This commentary is insightful. I recommend it to others.