Data on Donuts to Make You Go Nuts

Dunkin’ Donuts will roll out a loyalty program later this year.  (Shockingly, this will not be automatically included with residency in Massachusetts or Rhode Island.)

Anyone who has spent time in New England (or the Northeast in general) understands that this is an idea whose time is overdue.  Dunkin’ Brands’ Vice President of Global Consumer Engagement didn’t give away too much when chatting about the new program with Ad Exchanger, but this part was pretty interesting:

We look at the consumer base and there are a number of options for them to consider when making the decision on where to get their coffee, whether that’s in the morning or the afternoon. We wanted to not only reward them, but provide incentives, and ways to drive frequency and customer retention… [T]hat’s where every marketer wants to be — where they know exactly who is buying, when they are buying, what they are buying and what triggers they need to make them potentially buy more. We see it the evolution of driving a more comprehensive CRM opportunity with Dunkin’.

They may not know it, but it’s a direct copy from the 2012 Obama campaign: tap into an audience’s excitement, then gather as much data about what causes that excitement to translate into action.  Every cause has its champions, so ever would-be mob leader needs to think this way.

At least in theory, Dunkin’ Donuts will run a loyalty program that looks like a Presidential campaign.

Obama’s data may scale – but will his support?

Democrats are clicking their heels at the prospect of using the Obama 2012 list for the 2014 campaigns.  Fresh off a special election win in Massachusetts, the main concern seems to be how to scale the data from a national campaign down to a Congressional-level race:

That’s not to say, Democrats caution, that there’s nothing lost in the scaling process. Hiring a talented analyst doesn’t mean a campaign will be able to collect the immense trove of data — and update it over and over — the way the Obama campaign did. Not every Senate and congressional candidate will have the wherewithal, or the inclination, to test the effect of slightly varying messages on an experimental slice of the electorate.

But down-ballot campaigns also don’t need that level of data awareness in order to improve their performance in some material way. And if the Massachusetts special election was one case study in transferring data and analytics tools to a nonpresidential level, Democratic operatives say there’s plenty more where that came from.

There’s a big problem, though: there isn’t plenty more where that came from.  Barack Obama is no longer running.

Sure, he’ll “sign” emails – and despite tumbling approval ratings, that will mean a lot to a certain subset of voters.  Even so, starting in 2014, Democrats have to deal with a problem Republicans have  suffered since 1988: the specter of a popular and philosophically grounded President may hang over the election, but the candidates who fill his spot on the ballot won’t match his charisma.  Voters vote for candidates more than they vote for ideas.

And Barack Obama ain’t walkin’ through that door.



Thank the first responders, then enjoy the SPAM

Everyone has been standing with Boston and Massachusetts over the past week.  Finally, according to the Boston Globe, some political folks figured out a way to benefit from it.  The Democrats get the dubious honor of finishing first in the race to tastelessness:

Democratic National Committee chairwoman Debbie Wasserman Schultz was behind this tasteless tactic, sending out an email and tweet asking people to sign a supposed “thank you note” to the first responders. That would be nice except for the fact that in order to “sign” the note, you have to give the Democratic party your email account and ZIP code.

“We’ll collect every note we get and deliver them to Massachusetts Governor Deval Patrick and Boston Mayor Tom Men­ino so they can pass along your 
 Schultz writes.

A note to the brave police and firefighters?  How sweet.  All you have to do to “sign” it is yield enough information that the DNC can figure out who your Member of Congress is and a way to contact you.  More important, they can tag you as someone who cares about first responders.  Then, if they were so inclined during the next budget crisis and certainly during  the next campaign, the Democrats can reach out to you and tell you how voting for their person means safety, while voting for the Republican means a living in a post-apocalyptic war zone where warlords fight each other over guns and gasoline while neo-feudal serfs cower in terror.

That is, if the Democrats were interested in using first responders as political chess pieces, which they would never do.

Winning in ’14 means organizing in ’13

The battle lines in Washington, D.C. seem static: you have the Democrat-controlled Senate, the Republican-controlled House, and the President’s veto pen in a standoff.  The players are pretty much set on the chessboard, but no one is moving.  The legislative agenda hasn’t congealed yet; AdWeek even noticed the dearth of issue advertising.

Rest assured: America’s state capitals are making up for it.

Last week, Americans for Prosperity announced they will work to support Indiana Gov. Mike Pence’s proposed income tax cut.    Education reform debates have sprouted in Alabama and Ohio already this year.  Pennsylvania has considered privatizing its state-owned liquor stores; New Jersey is thinking about the same for its state lottery.

Heck, Rhode Island even named a state cartoon.  (Hint: It’s Family Guy.)

For candidates and issue groups looking toward elections in 2014 and 2016, these state-based issues and battles offer golden opportunities for organization.

After a grueling Presidential election and an apparently endless series of federal budget crises, voters may still be sick of hearing about national politics.  (A fair amount of folks who work in national politics are probably sick of hearing about it, too.)  The posturing plays out like predictable television dramas with familiar players.  Often, policy proposals are bogged down in process; and a hyper-partisan environment makes for fewer undetermined “swing votes.”

Meanwhile, state and local issues often involve ideas that haven’t been discussed over and over by talking heads on cable news.  Away from the D.C. spotlight, there is more room for Republicans and Democrats to seek common ground.  Perhaps more importantly, there is also more room for voter involvement.

By reaching out to voters and mobilizing them around state issues, campaigns and issue groups can begin building the volunteer and voter base needed for future ballot box victories.  Voters who take action now may self-identify crucial attributes such as issue preferences, favored mode of contact, and propensity for civic involvement.

Organizing around state-based policy discussions can yield information for a campaign manager that traditional attempts at voter identification might miss.  Turnout models may peg a hypothetical low-propensity voter as the type who would sit out mid-terms elections.  But when that voter responds to an online ad or a patch-through campaign, he or she has indicated an issue that moves him or her to action.  Further communication will yield more information.

The strategy of organizing around issues forces each voter to make a choice about whether that issue excites them or not.  By taking an action or refusing to do so, that voter shares a piece of information.  Enough of those pieces add up to a portrait of how to get that potential voter out to the polls.

President Barack Obama’s 2012 campaign has been widely – and rightfully – praised for its use of data to connect with and mobilize voters.  But these concepts are not new, nor are they particularly hard.  Careful tracking will give a campaign looking to win in 2014 a virtual “how-to” manual for speaking to the voters they need for victory.

Twitter predicts the Oscars… sort of

New Media Strategies predicted most of the major Oscar winners looking at social media data.  Brandwatch pulled a similar trick.

America has plenty of elections, from the crucially important annuals like the Oscars or the meaningless Presidential elections that we only bother with every four years.  In many of them, online networks and social media can predict results – winning candidates tend to be mentioned more on Twitter or liked more on Facebook.

While some will jump to the conclusion that online chatter will drive the support that pushes a candidate over the edge, that’s an over-simplistic reading of the situation.  Social media posts are tea leaves of human behavior, but not usually the initial driver.  It’s worth watching data trends and extrapolating results, but trying to create those data trends to ensure a specific outcome is a waste of time.  Daniel Day Lewis didn’t win an Oscar with social buzz, he won by making the legislative posturing surrounding the passage of the 13th Amendment interesting and engaging.  He didn’t even have to slay any vampires, so that was good too.  Similarly, online activity follows good political candidates, it doesn’t create them.

(Sidebar: What kind of a sick joke is it that Lincoln Motor Company is a subsidiary of Ford?)

If the correlation between online data and reality was more direct, according to Google we’d all have the flu by now.


Not just what it says, but where

Michael Turk had a great post on the center-right’s tech/data gap yesterday – but the best part was where he wrote it, in the American Spectator.

Spoiler alert: Turk warned that investing in new technology is not enough, that Republicans need smart people thinking about human behavior and voting patterns as well.  Good call: It’s not enough to figure out how people are interacting with a campaign, since most people in their right mind run away from political communication.  There’s an academic component in figuring out how to reach these people and keep them from running.  (Unless you use glue traps, of course, but there’s some questionable legality there.)

Ok, the right needs thinkers.  Where do they come from?  Political parties are good for resources, but not always innovation.  Remember that while much of the Obama infrastructure has been bequeathed unto the Democrat National Committee, it was the Obama campaign that built all the new toys.  Plus, if the eggheads don’t show immediate dividends, Republican candidates will wonder why the national party money that could be helping them win air wars is being spent to pay Lewis Skolnick.

The best spot for a bunch of data nerds is somewhere in the non-profit universe – whether it’s with an educational foundation like Heritage, an activist group like Americans for Prosperity or FreedomWorks, or a super PAC like American Crossroads/Crossroads GPS/Conservative Victory/Crossroads: The Next Generation.  With no donation limits, these groups can make a much better case to the big-ticket donors they’ll need to get the ball rolling.  Since the checks can be bigger, it’ll take fewer of them.

Conservative movement non-profits could be better positioned to start the process.  That makes The American Spectator a pretty good place to raise the issue.

Googlizing Campaigns

If you caught the tail end of the Roger Hedgecock show on Friday night, you may have heard me chatting with guest host Matt Lewis about the use of data in campaigns.

Much has been written in the past few weeks about the amazing things the Obama 2012 campaign did in identifying and turning out voters.  Just as much has been written about the Romney campaign’s failure to do the same thing, but it isn’t quite as fair.  There were many reasons Obama won, but the ability to take advantage of more channels of information to identify voters was a big part of it.

The private sector has been doing this for years.  For advertisers like Google and Yahoo! and e-commerce sites like Amazon, knowing what you do and  where you click online is their bread and butter.  It helps them put products in front of you that you’re more likely to buy, because they don’t make money if you don’t click.  Obama’s team was better at adapting those techniques to the campaign world.

What I didn’t get to talk about with Matt do to time constraints was the fact that Republicans can take a great deal of solace in the fact that these aren’t new magical spells being cast by technological wizards.  These are old hat tactics that can (and probably will) help Republicans with in the next campaign cycle.  For years, the advertising dollars have been moving toward personal advertising (like online ads) which can present content to an audience with much greater precision than mass advertising.

Romney adviser Stuart Stevens was ridiculed for saying that Mitt Romney ran less of a national campaign than Barack Obama, but he’s right, and Obama was right to do it.

The Googlization of Government

Rep. Tim Huelskamp has been banging the drum on a proposed Health and Human Services rule that would mandate insurance companies share patient data with the federal government.  The purpose of the program ostensibly noble – the administration wants to collect as much data on health care as possible to determine.  But Huelkamp correctly notes that data is not always secure.  Companies and governments lose personal data on customers and citizens periodically.

In a related story, Google revealed that the US government asks the search company for more user data than any other government on the planet.  In fact, there were more requests for Google data than there were wiretaps on phones last year.

While Google may look skeptically on the government requests for information, the HHS program sounds like something out of Google labs – aggregating data about users of the health care system to ensure better future outcomes.  Just as Google has multiple touch points where it meets its users (search, YouTube, Android, Gmail, etc.), so does the government.  What if they started connecting the dots?  We send tax returns in each year, so the IRS knows how much we make, where we live, whether we own or rent, what we do for a living.  On a state level, readily available voter registration data tells them how often we vote and may even give them a good idea how we would vote, based on primary voting history.  That doesn’t even get into people who participate in federal programs for medical help, student loans, social security, or public assistance.  And it doesn’t take into account the possibility of government looking elsewhere for data.  Today it’s Google, but a host of other companies are out there looking at what you but, what magazines you subscribe to, how often you gas up your car, and what TV shows you watch.

Eventually, other government agencies could follow the same model as HHS, expanding their data points on each citizen.  That’s when it could get really interesting, especially if some enterprising staffer in some agency realizes all the information that’s pouring in.  Imagine if the roadblocks between executive agencies came down, all the data was in one big pile?  The administration could be an even more voracious consumer of data, and use if to create detailed analyses of national trends, attitudes, and issues.  Here’s a video representation of how this might look:

A campaign or company wouldn’t use available data to recruit new customers or make life better for existing ones.  When I go to Amazon or Best Buy’s website, they look at what I’ve bought in the past and make recommendations; it’s simply good business.  An executive agency, which is supposed to strive for efficiency, would pick up on this trend as a way to streamline government services.  The difference, though, is that if you’re creeped out, you can always shop somewhere else.