Donald Trump needs to put his big data act together

By: Dhiraj Ahuja

Look at @RealDonaldTrump tweets and you’ll get to know that his campaign is more about insults and personal attacks than self-promotion (be it the picnic print of Hillary’s dress or her not being computer-savvy). His campaign seems brash with elements of reality TV rather than the typical ‘polish’ one associates US Presidential elections with – and Republican loyalists are rewarding it. He has a very good understanding of how the traditional 24/7 controversy or breaking-news driven American media works.

However, when it comes to digital marketing, he is months behind Hillary Clinton for she has been collecting data ever since she announced her candidacy around a year back. She has been serious about her digital campaign, a tool that President Obama used effectively for winning two terms in a row (outspending the competition with a US$ 47 million bill). On the other hand, Trump is complete contrast for he doesn’t seem to be doing much with data, calls digital marketing an overrated tool and plans ‘limited’ use of analytics. According to the reports, a few days back, Trump spent only one third of the money he invested in caps and T-shirts, on data. Republicans seem to be making the same mistake all over again – remember how Mitt Romney’s data effort started late, was limited in scope and finally crashed.

General elections are in November this year and therefore, it is too early to say about the delay in digital marketing hurting his prospects. One is sure though, Hillary has had clear and valuable head start and it matters. Nothing surprising, given the fact that Democrats have been leading for little more than a decade now and Republicans appear to be chasing them in this science of precision digital marketing data and online footprints. Besides, Hillary also has the advantage of the legendary Obama data team doing the job for her now. Based on the information derived from data, the ‘team’ has been tracking the supporters that visit her events and donate for the campaign, deciding on the type and volume of advertising, deployment of campaign staff and her ‘must-visit’ places. Not to forget the online content and mobile content voters consume, shopping choices etc. – everything is being tracked and analyzed using ‘cookies’ and a study of Facebook ‘Likes’.

Some say that Republicans have studied their mistakes, have invested heavily in data, hired data scientists and that Trump will go for a digital kill towards the end relying on his own data, Republican National Committee (RNC) data and that collected by Data Trust, a specialist organization funded by RNC. Whatever be the case, as of now, Trump and his team do not seem to be in a rush to catch up and this may prove fatal. Harvesting data is a must, one cannot be just sitting on it. It takes time, minimum six months or more, to tailor the data according to voter behavior and preferences and come up with meaningful predictive models. General elections are just five months away.

Trump seems to be having the momentum as of now (having ‘beaten’ Ted Cruz in the race despite the fact that Cruz used data analytics heavily) but that’s because of the great media coverage (PR at work) that equals US$ 2 billion in its advertisement value. Yes, he has had the data crunchers surprised, however according to the poll pundits, this approach of his worked just fine for the primaries wherein he was seeking Republican loyalists only. It may not augur well for him in the general elections this November as then, he would need to seek votes from all. And now is the time he would need his campaign to be ‘crowdfunded’ the most. He is at a disadvantage because he hasn’t built the database for crowdfunding over the year. Hillary has. Without targeted messaging, he is expected not to have a great idea of where he’d win for sure and where he wouldn’t, and where should he put his resources correspondingly.

Data driven behavioral micro-targeting powered by attitudinal information (and the algorithms developed by technology startups around campaign business) has the potential of transforming election campaigns into potential ninjas that deploy always-on voter feedback arsenal for hyper-local, regional and national contests. The thumb rule for smart and efficient poll strategy is simple - more data means more effective campaign or more bang for the buck. You cannot reach out everyone and therefore you need to identify and reach out to most persuadable people. Reaching out to those opposed backfires because it further reinforces their resolve to vote for your competition.

So, the questions that we do not seem to have an answer for as of now but would have post November results, are:
Without the predictive models in place, how would Team Trump identify the most persuadable and the ones who oppose him? In complete contrast to media consumption patterns and to the utter bafflement of big data experts, is the PR (traditional media) winning over PR 2.0 (digital media) this time? Or is it the anti-incumbency against the Democrats (they have been at the helm for 8 years) that’s giving him the slight edge he seems to be enjoying over Hillary currently? Last but not the least, would he retain this edge till the end and sharpen it even further (ad tech companies can work as a force multiplier and get him over the top) to trump Hillary or would he get trumped? Will Democrats pip the Republicans to the post yet again?

***
Dhiraj Ahuja has been a senior technology journalist with leading IT magazines and a PR practitioner for about two decades. He runs a digital campaign consultancy for political parties and candidates.