Big Data

The real big thing: data, data, data…

According to Statista, German companies alone invested more than 13 billion euro in advertising in the first half of 2015, which is equivalent to more than 160 euro per person living in the country. It’s a lot of money and it can achieve a lot if it is invested sensibly. Not every channel and certainly not every hype is the right choice in this regard.

Marketing by way of indiscriminate all-round distribution can never be as effective as that, which is pushed to the right recipients at exactly the right time by analysing and controlling all measures and channels. This may sound logical, but it still hasn’t fully permeated the marketing world.

The variety of options and (new) channels all too often results in hurried “me too” behaviour, zealous activism and ever more complicated measures with minimal results. Just because the competition is active on Twitter, for example, doesn’t necessarily mean that it makes sense to engage with one’s audience in that way. On the customer journey from mild curiosity to actually making a purchase, this individual channel may only play a very small role, if it plays a role at all.

Marketing departments are therefore advised to closely analyse this customer journey with all its touchpoints and control their marketing activities based on attribution models, which determine the actual added value of each advertising channel implemented. Such attribution programs can evaluate between 250,000 and 800,000 datapoints from a five-digit number of users and then bundle this information to deliver the right online ad within 30 to 50 milliseconds depending on a few user-specific parameters.

Knowledge and the amount of data grows exponentially with each operation. This is why the term “big data” is often used in this context. Retailers who quickly respond in the right way to the knowledge gained from big data can improve their margin by up to 60 percent, as McKinsey consultants have found. Attribution refers not only to the digital channels, but can be used to optimize budgets for all channels including online and offline channels (TV, catalogues, etc.) as well as mobile channels. And of course this is true even if only a few channels are used.

The attribution models, which have mostly been static models thus far, assign a percentage of each sale to each channel depending on its position in the customer journey. These models are increasingly being replaced by more advanced, dynamic models, which reflect complex realities. The technology behind this is very complicated and includes intelligent algorithms, statistical procedures and even the inclusion of cancelled purchases, but it provides scientifically sound knowledge to effectively address target audiences, achieve a higher level of customer loyalty and increase ROI. The trick here is that these attribution models are currently “learning” and can thus respond dynamically to shifts in the market or changing customer behaviour.

This dynamic method also explains why attribution is not a one-off project, but rather an ongoing challenge and constant undulation between measurements and the fine-tuning of budgets, strategies and channel mixtures.

Leave a Reply

Your email address will not be published. Required fields are marked *