2008 Election Recount Part 3 - The Death of Microtargeting?

Part 3 of my series recounting the 2008 election.  You can find Part 1 here and Part 2 here.

So did Obama's election really kill Microtargeting?  Perhaps not, but it definitely puts it in perspective as a tactic that is not a "must do" when it comes to elections.  I'd also argue that Team Obama was the first marketing program to actually deliver on the 1:1 marketing promise of the digital revolution.

Microtargeting was typically thought of as a holy grail when it came to political marketing - you know the NASCAR dads and the Walmart Women.  The Clintons' political strategist Mark Penn wrote a book in time for the election season called Microtrends.  It seemed that it would help decide the Presidency this time around, but the traditional microtargeting tactic didn't seem to help Obama win the Presidency; to me it seemed that the Obama team used more of a macro marketing view powered by people in the long tail.  Allow me to explain....

Back in my AT&T days I built statistical models for consumer as well as business marketing teams.  A few of these were of the profiling basis and others were of the responder type.  The profiling ones used regression to figure out the key characteristics of a customer base.  We then used the profiles for marketing purposes by varying the creatives and the messaging especially when we didn't have enough dollars.  However, the two fundamental flaws of this tactic are that you have modeling errors and that the variables that you can plug into the model are limited by the dollars you have and the people you have to crunch the numbers; basically there are almost unlimited amount of variables that you can plug in.

The political analogy looks like this.  In a central office a group of quant jocks examine polling data to see what the latest trends are and how the candidate polls by geographic groupings.  There are social trends (religion, guns, marriage, etc), economic trends, national security trends, and etc.  Based on the polling data you map out a messaging and targeting strategy for the intersections and then what the voter segments look like (voter profiles).  Then the quant jocks use their data to score lists of customers (voter registration files) and wham you have microsegments. 

This really isn't that much different from the private sector.  They key point here is that like the private sector there are centrally located analysts that push the marketing segments out to the rest of the team.  From what I can see and from what I experienced, Obama didn't operate with this model...here's what I think they did.

1. They obviously tracked their polling around macro groups of voters.  New voters.  Hispanics.  Blacks. Whites. Women. Democrats. Republicans, Independents.  Youth. Seniors.  Large important voting blocs.
2. If they cut lists and altered marketing based on these segments it was very subtle and I didn't catch it.  They obviously had the money to microtarget, but then again they had more than enough money to not worry about microtargeting.  Based on my 10 years at AT&T, 5 years in the brokerage world, and 10+ years in digital marketers only worry about microtargeting when they have scarce dollars and need to focus them in a few key groups or likely responders.
3. Team Obama bought a ton of TV and of course had the network roadblock.  Network TV advertising is expensive and untargeted (especially when compared with Digital).
4. Obama's digital advertising supported #3.  The banner ads were everywhere including Drudge and Townhall.  The ads were beautiful and smartly used Flash, but the messaging didn't vary much (join us, visit site, or find polling location).  They made extensive used of network ad buys and I do know they used retargeting pixels; retargeting pixels proves that they had a smart plan of funneling people through a process to continue the conversation on their own terms.  Perhaps the remarketing is where the customized messages occurred, but I couldn't be sure (I didn't make it a habit of clicking on opponent ads).
5. They maintained from the very beginning a 50 state strategy.
6. Team Obama made the most extensive use of the Long Tail of Marketing by turning the customer segmentation over to their people-powered campaign.  As I wrote above there are always errors in customer segmentation because you can never put in every variable into your model.  However, by empowering their supporters with tools from their website, the supporters were able to make their own personal customer segments.  Of course they didn't name these segments but when you get right down to it - if you speak with enough of your friends about their issues, then you can create your own messaging for them and you know how to speak to a customer segment of 1.
7. No other marketing campaigns today can actually speak with a customer segment of 1. I'd argue that Team Obama was the first marketing program to actually deliver on the 1:1 marketing promise of the digital revolution.  BTW - it wasn't that our campaign didn't offer the same basic tools, we just lacked the money and people resources.
8. Of course they could have data mined the information to look at people that didn't register, unsure, etc and their army of supporters were only too happy to provide the information for use - they had an army of data collectors more powerful than any list of polling samples.

With Team Obama's unlimited funding sources and the huge list of volunteers and staff, they were able to throw the old playbook out the window.  They didn't need to use a micro-targeting approach. 

Obama was able to focus on macro voter blocs and then they provided the tools to their army of volunteers who attacked the Long Tail of voters on their behalf. 

Did they kill Microtargeting?  Probably not, but when you have unlimited funds and a product perceived to be just what people wanted, you don't need to worry so much about Microtrends.  You can actually deliver on the 1:1 marketing promise of the digital marketing revolution.

PardonMyFrench,

Eric