Well, this unsolicited letter from Omniture (now part of Adobe)  had me scratching my head, since our company specializes in driving Facebook traffic.  Do a Google search on facebook advertising and see where we rank. Also, we have 4,920 fans on Facebook vs Omniture’s 3,410.

What is presumptuous about saying where our traffic comes from is that it’s not possible for Omniture to measure our ad traffic or traffic from various Facebook pages that we have a voice. The old world of web analytics assumed you owned the audience– they hung out on your website. But in modern times, that audience may choose to interact on Facebook with your brand or between friends and perhaps not even set foot on your website. They could be on their phones, in a coffee shop talking, watching TV together, or doing other word-of-mouth things that a pixel cannot capture.

Thus, even if Omniture could magically peek into our Google Analytics (and to which they’d see that 3-10% of our traffic is from Facebook), it would still be missing the picture.  In the world of PPC, there are keyword research tools where you can estimate how much someone is spending on PPC and break it down by keyword.  In Facebook, that’s just not possible because of the targeting options.  I could run ads targeting boys 13-17 that like to play football in the US– and unless you fit that criteria exactly, you just won’t ever see my ads. So to pretend to know how much traffic we get on Facebook is somewhat silly, don’t you think?

Are you also mistakenly assuming that it’s a good thing to rip users away from facebook and into your website?

Because of Facebook’s graph API, it’s now possible to measure your traffic in a user’s news feed (home page), where they spend about 50% of their time. That’s right, most of those likes and comments that may appear as if they are occurring on your Facebook page is actually happening in the news feed.  And though you can pixel a custom tab on your page, you cannot tag your wall, nor can you tag the news feed, which is an order of magnitude larger in traffic than what you can track via only the old way of doing things.

How much of your conversation are you missing by not listening in the right places?

P.S– Sorry, Omniture.  No hard feelings. I get a ton of unsolicited sales letters such as this one, so I randomly picked one to showcase.  If you’d like to write a counter response on how to properly measure the value of a fan, even if they don’t always come to the website (which is what your white paper insists that clients do), I’ll happily post it.

Read the MediaPost article, hot off the presses.  Omniture (now owned by Adobe) will have the ability to buy ads via SearchCenter, their PPC management tool, by the end of 2010.  That’s over 3 years behind BlitzLocal– an eternity in web advertising, so it will be interesting to see what product is in place by then.  I will want to know if Omniture has plans yet to integrate Facebook into SearchCenter such that it’s treated like another search engine or whether it operates orthogonally (since social media plays in the the AID of the AIDA funnel). 

It will be interesting to see how large agencies and brands adapt, given their penchant for wanting simple interfaces to manage campaigns, as opposed to performance marketing on Facebook.  The world of buying Facebook ads is significantly different than traditional keyword advertising– much closer to the content network, but still it’s own animal.

There is also the issue of attribution, since social media adds additional touchpoints earlier in the funnel.  On Facebook, people find out about brands and develop trust from seeing what their friends like and endorse.  To be able to now see those social touchpoints later convert on Google search will pose an interesting challenge to advertisers that have to now allocate budget between social and search.  This doesn’t even yet address the issue of brand bidding, which also messes up the perfect dream of automated bidding– to be able to hide behind the black box math that is actually just last click attribution.

Even when Facebook does introduce their own conversion pixel later this year, it doesn’t address the issue of how to allocate credit when a user has multiple clicks in their clickstream prior to a conversion.  Neither last click nor first click nor average click is the correct answer.

I believe in a multi-channel environment, the advertisers who rely upon having smart in-house strategists will win over those who decide to shell out cash to buy the most expensive software.  Online marketing, and marketing in general, is increasingly becoming multi-channel, which will scare some into hiding behind software, while motivating others into arbitraging out the profits amidst the confusion. 

We’ve been at this 3 years with brands large and small, having seen and experienced the problems that the big players have yet to even encounter.  What do you think will happen?