In an era where every marketing spend is carefully considered, marketing teams are constantly searching for the best way to justify marketing activities and costs. In a landscape with changing technology, legislation and increasing competition and expectations, there is one technique that should receive more attention and is a critical success factor for marketing strategies: data modelling, our modern day crystal ball.  

Data modelling extends beyond traditional approaches to analytics and enables marketing teams to take a predictive and measurable approach to planning and budgeting for 2024 and beyond. And yes, that is exciting and requires a change to working methods, but more about that later. 

One of the greatest values of data modelling is its ability to gather predictive insights. We apply algorithms and models to our past data to map future trends and patterns. This enables marketing teams to make decisions based on data-driven insights rather than gut feelings. It finally gives an objective method to map out your marketing investments.  

Another advantage of data modelling is the benchmarking possibilities and competitive analysis. In a world where companies compete for the attention of the same target group, it is essential to understand where your organisation stands compared to the competition. But when we barely have time to examine our own performance, we certainly don't have time to do that for our competitors.  

Data modelling takes manual work off your hands and once a model has been developed, it is there for you to use whenever you need it. It is possible to compare marketing performance with sector averages and competitors (benchmarking). 

Data modelling is nothing new. Recycled ideas? A little bit. It is the (digital) marketers, who are waking up now. Sectors such as logistics, retail and healthcare are light years ahead of us. Data modelling is taking off, helped by (Gen) AI, but it originated decades ago. Today we still use the same models as when it was first introduced. We see that these proven models have been developed and that usability is improving.

Models have now been developed specifically for marketing applications. For example, Robyn (Meta) and LightweightMMM (Alphabet) are two models for Marketing Mix Modelling that use different techniques to predict marketing effectiveness. These developments ensure that these techniques are part of the marketing scene.

The different forms of data modelling in the marketing context often have the same starting point. Most methods analyse data from the past and use insights to 'predict' the future. Models can analyse data quickly and find connections that would take much more time and energy using 'human capabilities alone'. Examples of marketing applications include: 

• Modelling customer behaviour: understanding behaviour to recognise purchasing patterns, churn behaviour, or cross-sell opportunities 

• Customer Lifetime Value: insight into customer types that are valuable or not 

• Marketing Mix Modelling: insight into the effectiveness of marketing channels and how advertising budget can be spent effectively 

• Sentiment analysis: insight into behaviour that contributes to positive or negative segment 

• Segmentation/personalisation: insight into the different customer groups and how they can best be served 

Although data modelling is becoming more and more accessible, it is not a magic box where you can just put an Excel file in, and it then conjures up the most valuable insights. 

First, organisations need a few years of historic data, depending on the application. This is the power supply for data modelling; Without data, there are no insights. Fortunately, today, many organisations have been measuring and storing data in a broad sense for a few years. Sometimes there is still work to be done, cleaning and bringing together multiple data sources, but a lot of organisations have already got enough data to get started. Important to remember this is not only for multinationals. 

Next, applying data modelling is of course a skilled profession, you need the right people. Data scientists can apply the right models with structured data and fine-tune them in such a way that the model is able to retrieve the insights that are relevant to the organisation. If data is still disorganised and unstructured, help from a data engineer and analyst is necessary. 

Organisations have to learn how to deal with insights using algorithms and models. Today, organisations have a lot of confidence in subjective and often biased insights, and less in objective models. And that makes sense, because we have less instinct and gut feelings about the emergence of the insights from models. In order to work with models, we must learn to rely on computer-controlled advice.

It is therefore important to recognise that joining the data modelling revolution is not a one-off decision. Explore the opportunities offered by modelling in 2024. The world of data analysis is constantly evolving, and it is wise to keep up with it. While we may not have all the answers by 2024, it's essential that organisations incorporate data modelling and data-driven approaches into their future plans. Use your 2024 results as materials for comparison with your gut led decision making, so that you have crossover time to build faith in them. Not only for yourself, but also for your colleagues. A striking insight from the aforementioned Gartner study is that CMOs sometimes have to scale down analytics teams because they have not been able to integrate the insights into the organisation. So, give the process time.

Data modelling plays a crucial role in accounting for marketing activities and costs, providing predictive insights and competitive advantage. As we prepare for 2024, we need to understand that the power of data modelling allows us to look not only at what activities have delivered, but also at where the future potential lies. And what it takes to harness that potential. Data modelling is at the heart of modern marketing strategies and will be a determining factor for growth in the coming years. Stop gambling, it's time for objectivity.

This opinion piece was written by Lars van Tulden. In his role as a data consultant at iO, he helps customers with all kinds of data issues.