“For example, an average camera buyer takes 30 days and clicks on approximately 3 sponsored links from the beginning of researching cameras to actually purchasing one.”

OK, but what else did that average buyer do that was not trackable? Talk to friends? Get referred to a friend of a friend? See a magazine ad? See people walking around using a certain camera? Visit a few stores?

And what about if this buyer uses multiple devices to do the research? How would you track that?

Let's say I read a positive review on a photography blog, pick up a cookie from that blog, and then decide to hold off on buying for some reason, then my friend shows up with that camera weeks later, I try it, I'm hooked and I buy the camera online without clicking on a paid link on a site that is able to track that cookie … well, how much should the photography blog get?

You're doing a lot of measuring and tracking but if you're honest you must realize that you're really no better off than the 'traditional' advertiser that accepts the old saw: I know I'm wasting half of my budget, but I just don't know which half.

The fact is that many buying decisions are made in a rather messy human sort of way. Just look at your own habits and those of the people around you. I have read positive reviews for products that I have purchased, but not for the reasons that I read about in the review. I'm not trying to be pedantic, I just happen to think that's how the world works.

If there is any light at the end of this particular tunnel it could be the holy grail of social network centric advertising with the caveat that we don't yet have any clear idea of what that will look like.

Most advertisers made their first contact with converters months prior to the conversion. Figure 1 shows the median number of media touch points at different times in users’ conversion histories. Unsurprisingly, the amount of advertising exposure is quite high in the last few days before a conversion. Consumers amass a median of 5.5 ad events in the final 48 hours before conversion alone.
But this frequency amounts to less than a third of the total media events delivered over three months. Half of the total engagements with converters occurred between seven and sixty days prior to conversion, and after a full 90-day look back, converters experienced a median of 18.5 events.

Here is the link to the .pdf of the study: http://bit.ly/aXuY6N

Google (and other CTR sellers) have understood the argument you made and have built empires around harnessing the last click, and showing advertisers “look I made the sale for you!” – it is something that I like to call the “Google Tax,” akin to giving the man/woman who rings you up at Target all the credit for selling the item, and discounting everything else in the supply chain.

But here is the rub, the internet is/has become decentralized. I suspect that our patterns of finding content and information on the web are changing and that soon peer-to-peer recommendations will become a more powerful form of spreading content – and that includes Marketing content. I think that this will threaten the status quo on the web – especially search. I am beginning to believe that a person can get a personalized content delivery experience without being dependent on a search. How does that effect marketing? In everyway…

I am new to your blog. This is great stuff, and an interesting debate.

The only comment to add is the following – As in all things in life, the funnel is part science, part art. While it appears most, it not all people here are steeped in science of “funnel analytics,” what appears to be missing is the art of communicating value propositions up the food chain at client marketers.

Those of us in Publishing/Networks, Agencies, or Digital Marketing at “Widgets R Us Co.”, all speak the same language, and understand the opportunities and immediate challenges in digital marketing. But does the CMO to the same level of detail? Or his/her CEO? Let alone their B.O.D.s’ and shareholders? The point being this – there absolutely needs to be new disruptive technology which helps marketers distribute advertising funds appropriately throughout the pipeline (aka Funnel) and provide them the ability the manage it. And quite honestly, it should be funded through a JV of marketers, agencies and publishers.

At the end of the day, branding, research/thought leadership, couponing/last click all have their place in the marketing continuum. I think we all agree on that. It’s how much of what, and when do we need to make adjustments, and to what degree, is at issue. Demystifying funnel analytics into straight forward, sales and marketing strategies, coupled with good analytics >>> will lead to better engagement at top management at clients >>> and in turn will lead to more dollars flowing into the digital economy >>> which will help fuel innovation, supported by the 3 constituency groups.

Final thought for those who “hate” advertising, I have 1 question for you — You mean you didn’t let out a little chuckle, let alone a smile of amusement when Betty White got tattooed in that Snickers commercial?

‘Nuff said!

Content sites who reel people into being first time clients, if this model can be revised from “% of the last sale from the last click” to “% of the customer lifetime”, should be able to earn major affiliate revenue.

* Asked a bunch of friends about what laptop to go for.

* Surfed through a couple of laptop review websites.

* Went to CircuitCity and BestBuy and checked through a bunch of laptops.

* Found 2 friends who wanted to buy a laptop too (starting of first year of college – this was easy.) So went and negotiated a better price from CircuitCity to buy 3 laptops together.

Probably 10 people helped me make the decision. But only CircuitCity got the money directly from me.

Online buying isn't that different than offline buying.

My money goes to CircuitCity. Part of CircuitCity's money goes to Toshiba (the brand of laptop I'd bought.) And then its upto Toshiba to spend their money on advertising and PR to make sure people know about their products.

The funnel is working. Its just not a simple funnel because not all products have equal opportunity of being bought.

The end vendor pays monies to interest harvesters. The money from interest harvesters trickles down to interest generators*

*unless they're reviewers. There has always been a gap between vendors paying money to reviewers. Because thats bribing. But companies like JD Power has cracked that code. Which is really the only way of packaging things up and make sure the reviewers get paid their worth.

The funnel – imho – is not really that broken. But not all levels of the funnel adhere to the same metrics and tracking devices.

Any misallocation that is happening is happening because of company planning error – and not because the funnel is not efficient.

The one aspect of this that you have not covered, which is one of the powerful factors prohibiting innovation in this space, is the power of the voucher/coupon code sites.

I have spent a huge amount of time talking to affiliate networks about endorsing solutions like the ones provided by Convertro and Tagman. The reasons this is not being done are:

The cost is charged to the merchant/advertiser, and there is little incentive for them to invest in this disruptive technology. By adopting multi-attribution, you completely topple not only the power, but the business model basics behind how voucher/coupon and cashback sites work. They rely on their ability to know precisely how much the customer will save/make on a purchase, because they can guarantee they are the last click. The second you bring in multi-attribution, they can no longer accurately tell their clients what they will save/make, and their business is undermined.

Unfortunately, these sites command an immense amount of power in the affiliate marketing industry. Love them or hate them (and most merchants hate them but can't escape them), these sites drive enormous amounts of sales. If they refuse to work with cashback/voucher/coupon sites, their short-term sales take a significant dive, and even more infuriatingly, their competitors will simply take their place on these sites and earn the sales they would otherwise have gotten. I have spoken with many of the top global voucher/coupon/cashback sites, and they are very adamant about their strategy when merchants/advertisers stop working with them.

So, merchants have this dilemma: do I invest in a technology that will jeopardise my relationship with the sites that drive 95% of my affiliate sales, just to ensure that perhaps in the longer-term, more content sites will write about my product, maybe.

This is what we are fighting against to win this battle, which I dearly want to win. But to win means convincing merchants that the short term hit to their revenues will be more than compensated in the long term by more revenues caused by a widening of a better-funded top-end of the funnel. For this, we need metrics – there is some incredible work being done in the UK by very clever agencies. I'm working with them to pull together a case that justifies this leap of faith on the part of the merchants. Hopefully once we have the metrics, and some great case-studies, we can convince more merchants to make the investment in technology and resource, and start sharing the rewards across all those who participate in creating an intent to buy.

I was hoping you, with your background in philosophy, aspire to stay clear of shallow opinions. To dismiss Jaron who is the most significant and honest internet thinker today is just intellectually weak. If you are not prepared to debate his assertions, I have a hunch your are not prepared to engage the Internet culture with clarity.

She didn't do it intentionally but she is also blinded by her circumstance. I understand you have to defend her.

The “paradox of quality site visitors” is that a domain-typo-squatter's site which induces people immediately to click away monetizes (back in 2006) at 10x the per monthly-unique-visitor rate as does a content-rich community site.

Why? Because CPC as a model was rewarding the “next click” (if not the “last click”) and ignoring the amazing demand generation that was going on in the hobbyist community site.

As far as your reference to “you people” I should note that I'm not a fan of advertising, in general. In fact, I loathe it.

I wish there were more outlets like Kickstarter and 20×200 that are disrupting the way creatives can earn a living. Absent those platforms though, the unfortunate reality is that advertising is the easiest way to subsidize content in a world where consumers don't like to pay for it.

I believe in time, we will start to see a paradigm shift from where we are today (consumers getting hit over the head with a message that they didn't ask to see) to one where ads can actually serve some utility and be part of the user experience – with proper compensation being distributed appropriately. Perhaps that view is idealistic, though I am hopeful.

There is a macro evolution happening which is going from a “view thru” society to a “measured engagement” society. Measured engagement can include “clicks, sales, leads, downloads” or whatever measurable proxy for performance.

Because the commercial Internet is less than 20 years old, we're still in the early innings of this game due to the legacy measurement systems that were built. Because the entire advertising ecosystem was built around legacy systems (at the time, didn't feel like legacy), it's going to take some time to change this.

People have been mentioning convertro in the comments but also check out companies like joviandata and visualiq.

The macro question: you can have cutting edge measurement that moves away from last click attribution, but when will madison ave understand the advertising ecosystem when it moves to near-real-time (or even real-time)? Is Comscore relevant anymore?

I agree with you wholeheartedly that there is incredible interest and intent being generated by publishers but I don't believe it's due to the advertising that's on their site (banners, sponsor or otherwise). I believe it's because of their content (your CNET example is a perfect illustration).

That said, I think there are two issues here. 1) attributing top-of-funnel advertising to bottom funnel performance & 2) attributing content or context to ad performance.

The latter issue is one that publishers will have to grapple with themselves (good luck finding and advertiser or agency that's going to happily give away her dollar). I think there is tremendous opportunity here though. I'd say Rubicon has a head-start here.

The first issue is one a small number of really smart advertisers are already tackling e.g. buying banners to drive up brand (cheap) keyword inventory. However it's really a cross-media attribution issue more than a siloed web issue.

The earlier poster is right, Microsoft was leading this research a few years ago however in recently speaking with Google's display and mobile teams I would say they've lapped them at the moment.

What's great is that we're already the most measured and attributable medium yet we're far from satisfied. It will be fascinating to see what and where our real value rests and how that impacts other channels in the coming years. A bounty of opportunity awaits.

Sorry for the links back to my site (I hate that) but post to post communication is one leg up from the social media.

Of course the problem of getting paid for authorship goes way beyond just online (people do a lot of talking and reading before buying something like a car…most of the time, they aren't paying anyone for most of the 'research' that went into their decision)…but my orig. point was just that it seems like a lot of people are really starting to think about this sort of problem (from all different angles) and so it's interesting to see the different angles and watch how everyone tries to work it out

Can anyone point me in the direction of information that details what the big ad serving platforms (DART, Atlas, Mediaplex) are doing in the attribution space? I've seen some mention of Atlas, but no links.