It is time to put ideas first and data second, argues Yannis Zachos, head of strategy at Havas Media International.

It is not hard to understand why data is still the killer buzzword in company boardrooms and marketing departments. Data-driven marketing holds the promise of accountable investment, increased effectiveness and reduced risk.

The power brands that were born out of the digital age are leading by example. Running on data-hungry algorithms, they are always testing and constantly optimising. Every movie we watch on Netflix, every song we listen to on Spotify, every search we do on Google generates data that improves the algorithms and subsequently the user experience.

These brands represent the golden standard when it comes to data driven decisions. However, our troubles start when we take their successful models and try to apply them to non-native digital brands.

It is great to have big aspirations but it also helps to be pragmatic.

Incremental versus transformational

Data solutions are often seen to have the ability to generate insights that if acted upon, can be transformational for brands. Whilst I do not dispute this possibility, I’m still to encounter an instance where this has happened.

Whether it’s social listening, search data, sales or other transactional data, it is extremely rare to discover a fundamental insight that has been missed by the traditional quant or qual research.

On the contrary, I’ve had many instances where data insights were used to achieve incremental results.

“It’s time we stop worrying about building the latest, state of the art data platforms and instead focus on what we need to do to connect brands with consumers”

A recruitment client increased its lead generation by 15% when shifting offline activity to midweek placements. It did this after noticing that Wednesdays consistently delivered high online engagement. Turns out people are more likely to look for a job midweek, hence they are more perceptive to relevant messages.

For a men’s retailer, it was the realisation that about one fifth of its customers were female that led to a joint gifting campaign with the sister female brand. As a result, sales grew by 30% across the board.

When a luxury e-commerce fashion brand turned its global sales data into live trends for prospective customers, the campaign ROI increased by a tenfold. Interestingly enough, people care more about what their peers are buying than the latest magazines trends.

None of the above cases required vast data analysis or complicated data platforms. These brands did not set off with a manifesto to deploy data marketing. Instead, they started by asking questions and conducting interesting hypotheses.

People not numbers

Interpreting causality through data can be crucial for businesses but so is understanding motivations and behaviours. By aspiring to create a single customer view or to improve the various imperfect attribution models we are running the risk of downgrading the importance of cultural observations and human truths.

Data cannot unveil anthropological insights like for instance that men need an activity together to make and keep a bond as opposed to women who can maintain friendships over the phone. Or that men tend to make their deepest friends through periods of intense engagement, like school or military service or sports.

A beer brand can use data to identify the best moments to engage with beer drinkers but that alone is not enough. Because this is exactly what every other beer brand will be doing. However, if their strategy is to position the brand as the essential part  of male bonding that would unleash a more creative and diversified data strategy. Instead of trying to use data to win the customer at the “moment of truth” better aim to win the customer’s heart way before purchase.

It’s time we stop worrying about how we can build the latest, state of the art data platforms and instead focus on what we need to do to connect brands with consumers in meaningful and creative ways. Ideas first, data second.