Many media reports about data-driven campaigning are sensationalized. We all remember the story about Cambridge Analytica, which was foremost good marketing for their own firm. And many other accounts about the wonders of data in elections is told by companies who sell such data.

Empirical evidence shows a more nuanced reality. Sure, most campaigns collect data, but how much they collect, what kind, and how they use it varies widely across the globe. Not in every country is the same data (legally) available. And not all parties have the same financial capabilities, personnel and willingness to mine and drill it.

An entirely different question is the quality of the data. I have always found it flabbergasting how blindly many campaigners believe social media data, when there is in fact plenty of reason to doubt it.

In my experience, there is also absolutely no need to get data behind the back of voters. They are very willing to share their opinions. We just need to listen.

That said, it’s clear that artificial intelligence offers tremendous possibilities to campaigns. One of the things AI is exceptionally good at is analyzing huge amounts of data. We can combine various sorts of data (survey data, consumer data, past political behavior), build sophisticated models and track in real time. It is indeed promising — if campaigns are willing to experiment.

But there is an even more fundamental point I want to be clear about: data should inform strategy, not replace it. Because the goal is not to describe the universe, it’s to win the election. Data must inform and guide action steps leading towards that goal.

Dr. Perron’s Roadmap to Victory® is my registered trademark in the U.S. It describes the three-step-process for data-driven campaigns that I have used multiple times to guide candidates and parties towards electoral victory.

But I have often observed that data becomes an end in itself or that people struggle to translate data into action steps. Some clients, after my presentation, ask for more data, more segmentation, more modeling. They are usually the ones that get lost in the data and don’t act.

My most successful clients, on the other hand, say “based on your analysis, tell me the five things I have to do.”

And then they do them. And they keep winning.

PS: Check out my video on YouTube where I discuss data and strategy using examples from the world of tourism and my check-in experiences in various hotels.