Red Queens and Red States
Today I have spent 6 hours on a flight from New York to San Francisco and excitedly blasted through the entirety of a new book, The Victory Lab. To an avid political buff who does market research work (a narrowly defined subsegment), it is a fantastic read which I would highly recommend. It really does highlight both how far behind political campaigns were on the analytics revolution compared to real marketers, and how far political campaigns have come in the past ten years or so. It’s also surprisingly heartwarming to learn that the best method of influencing voters really is having a dedicated volunteer have a conversation with them. I suspect that in the long run, the late 20th-century political campaigns focused on mass-media “air wars” will look like an aberration.
This book also has strengthened my belief that the whole Analytics/Big Data trend is actually close to peaking in the returns on investment. The gains to political campaigns from better microtargeting, enhanced A/B testing, and improved voter tracking will continue to matter – but they will be incremental compared to having the base technology required to ID voters on an individual basis, track support and touches, and maintain a centralized and (very relatively) clean database of those entities. It will likely in politics, as in business, continue to suck up greater and greater investment while delivering smaller and smaller incremental returns. For all we know, analytics peaked in ROI yesterday, or three technological generations ago. But of course you can’t afford to fall behind in the Red Queen’s race!
It also suggests to me that while operational excellence is essential to being competitive in the marketplace, the way that you can truly excel in the marketplace is through having the best product. Whether in business or in politics, you can’t succeed with fantastic logistics if your product is shit. Go ahead and ask Apple’s competitors whether Apple is kicking their butts due to Apple’s mastery of the supply chain. Or ask John McCain if he lost because his campaign didn’t embrace A/B testing to optimize clickthroughs.
Finally, while this is notaddressed in the book, I am skeptical about the relevance of microtargeted appeals to tomorrow’s voter. For low and medium-information voters, social media is increasingly the primary channel of political knowledge. Combined with the fact that everything is now taped or might be taped, that makes it extremely difficult and probably futile for campaigns to imagine they can tailor their message to specific voters. You might want suburban Bucks County voters to hear about your tax cut plans rather than how you’ll ban birth control, but you have no control over whether that goes viral. It might just be that all the whole mechanism of micro-targeted persuasion communications is totally laughably irrelevant to the way voters actual get and process their information, and all these campaign practitioners are just going nuts on spurious correlations.