home > articles > 

MICRO-TARGETING IN 2010?
by Nick B. Fontanilla, Ph.D. (Word Document)

Consider a presidential candidate who can identify who among the voters, identified by name, are undecided and open to a better political proposition. Consider further that this candidate has a database of undecided voters, with contact details, that is big enough to influence the outcome of the election. Then consider this candidate who knows exactly the issues these named undecided voters consider important to make a decision favorable to the candidate.

Unlikely? Well, this methodology is what Ken Strasma, a Washington-Based consultant is currently doing for a presidential hopeful in the U.S. democratic nomination. Labeled micro-targeting, this was also the approach that some candidates employed in the previous U.S. elections with amazing results. This consultant-strategist claims that it is possible to target the most powerful message to each undecided voter, focus “get-out-to-vote” efforts on the voters most likely to actually support the candidate, and boost the returns from fundraising prospecting by mailing or calling those most likely to contribute if asked, and, most significantly, finding brand new donors, rather than just re-soliciting the same pool of know political donors.

Micro-targeting is not a guarantee for political success. Nothing does. However, the political candidate gets the best value for money spent. And only what is needed to get the message across. Every centavo is spent for the right purpose. “Huge amounts of time and money are spent trying to persuade people who have already made up their minds, to turn out voters who don’t support our candidates, and to solicit contributions from people who will never donate,” says Ken Strasma.

In last year’s Philippine election, some senatorial candidates spent so much of their resources to communication programs that reached out to everyone including voters who have already made up their minds and no longer open to persuasion, and even non-voters and non-contributors. They were delivering messages that were not coming across simply because they were designed for everyone or for a regionally segmented audience. The results were financially devastating. I am sure that those pockets really hurt.

What is micro-targeting? Micro-targeting is a methodology that predicts (with a definable level of precision) how voters a candidate is not able to reach would have responded if that candidate had been able to profile them. Strasma reveals that the idea behind micro-targeting is to look at what we know about voters we were able to ID, and to use that information to build a statistical model to predict how voters we didn't reach would have IDed if we had reached them. He combines voter ID information from phone or door-to-door canvasses with detailed demographic and marketing information in order to build the statistical models.

Using his statistical model, Strasma identified 34,208,571 Democratic GOTV target voters (including 7,368,609 voters in strongly Republican rural and exurban areas), 23,616,066 likely “undecideds’, 44,646,346 drop-off voters, strength of voter support for or opposition to more than 35 different issues, level of motivation among GOP and Democratic base voters. Starting in 2006, he was able to apply micro-targeting scores to more than 95 million voters, a knowledge that translates to a definite advantage and cost-effective campaign.

  Sounds corporate? Indeed it is. Micro-targeting uses best business practices including methodologies that financial institutions, retailers and government agencies currently use. These best practices include credit scoring, cross-selling, up-selling, insurance fraud detection, money laundering, traffic management, crime prevention and one-to-one marketing. These practices are all based on the principle of risk management.

Jack Noonan, CEO of SPSS, Inc., a leading global provider of predictive analytics software and solutions, wrote: “Not long ago, election campaigns were run on simple grassroots efforts dependent on cold-calling and randomly distributing flyers to the public without strategy. The reliance on individual voter data was very thin, and only gathered from the public registrar. Fast forward to 2008 when large political organizations to local campaign strategists are analyzing vast databases on voter files, election returns, and demographic data through the use of Predictive Analytics software. This keen analysis enables organizations to target likely supporters, persuade undecided voters, and help ensure citizens to get out to vote on Election Day.”

Noonan observes that campaign strategists now deploy this data to better persuade voters by identifying and communicating the right message on key voter issues, such as education, healthcare or immigration, through targeted and personalized direct mailings, phone calls, or door-to-door canvassing.

Can this be done in the Philippines and help candidates manage the risks of a political campaign? I believe so. First, we have the tools and the talent to apply the same methodologies. My company does this for private companies. So does a few other consultants. Second, while we do not have an efficient ID system or a database of voters’ profile, we have more than enough government databases or databases that are for sale. When compiled properly, they will form a formidable list that can be the subject of a statistical model deployment. Third, the cost of mounting a political campaign is getting so high that it makes sense to be scientific and purposeful. Atty. Alex Lacson of the “12 Little Things Every Filipino Should Do” fame once told me that he knows of three politicians in his hometown who lost everything (ending up penniless) after losing thrice.

Can this methodology stand the encroachment of politics Philippines style. I believe so too. In 2004, The Asia-Pacific Centre for Research (Acre), Inc. conducted many political surveys covering congressional districts and cities. Across all those surveys, the proportion of ‘undecideds’ a month before the election, which was close to 50%, significantly decreased a week before the election. Doing the same surveys in the same areas last year, we observed that the proportion of ‘undecideds’ was rather low as compared to the same time reference in 2004. Moreover, the proportion hardly changed until the week before the election. If this is a trend, it only means that voters make up their minds way before the election and, when they do, they are no longer open to persuasion.

There are just three things that a candidate needs to do micro-targeting. One, knowledge of predictive analytics or a political strategist armed with that knowledge, resourcefulness to acquire and compile databases from government agencies and associations, and third, openness to a new paradigm in managing the risks of a political campaign. If there is something really useful that came out of this side-splitting U.S democratic presidential nomination, it is micro-targeting.