Infographic

Flat vs. Rich Data Model for Your Marketing Automation

Flat vs. Rich Data Model for Your Marketing Automation

Choosing a marketing automation solution can be quite difficult and you need to take various criteria into account for this. One important factor is the data model. Some solutions work with a flat data model, i.e., a flat table database while others work with a rich data model, or a hub and spoke model.  

In this article, illustrated with specific cases and infographic below, we explain what you can do with each of these models. Everything depends on your marketing data modeling needs, the scenarios, the level of personalization… 

What’s the Difference Between a Flat and Rich Data Model?

What is flat data model? 

A flat data model is the simplest data model. It simply lists all the data in a single table, consisting of columns and rows. It is nothing more than an Excel file. You are thus limited by the number of columns you defined from the outset and you will have to think very carefully what you want to include in it. And there’s the rub, because your data will not be very adaptive. 

What is rich data model? 

A rich data model is a relational model which sorts data into tables, also known as relations, each of which consists of columns and rows. The rich model allows you to develop a more detailed customer segmentation, meaning more far-reaching and accurate personalization. The more information you have about your customers, the better you’ll know what to pitch and when. 

What’s the Difference Between a Flat and Rich Data Model?

Data Modeling for Marketing Automation

Marketing Automation is the engine of the customer relationship while the data is the fuel. And of course, once you have the data, you still need to know how to use it all! 

The simpler and the more basic your needs are, the better a flat data model will be for you. But if you prefer to communicate one-on-one with your customer, utilizing all his behavioral data (purchases, web visits, in store visits…), then you’ll need a rich data model. 

A flat model allows you to develop a cross-selling strategy based on product recommendations. But it limits your options as you will only be able to recommend one single product based on single interaction, often the most recent one. A rich model, however, lets you recommend as many products as you want, based on all the interactions during the customer journey. You can thus maximize your chances of suggesting the right product, which will please your customer. And this will only increase your customer knowledge and your ROI! 

Examples of Data Modeling in Marketing Automation

Your marketing automation model will depend on the industry you are working in and on the data sources you want to store and integrate. Retail data model can be very rich for example as you collect data for many sources (offline and online). In the examples below, we will show you why data modeling is important. 

Examples of Data Modeling in retail  

Think of a family with children. You want to offer a gift to each parent whose child is celebrating a birthday. You’ll need to use the parent’s data (“Hello Mr Jones, …) and his child’s data (“… your daughter Jane is celebrating her birthday today”). A parent can have several children, and you have no way of predicting how many. If you were thinking of offering all of them a birthday gift, then leave the Excel sheet for what it is.  

Good to know: the flat table model does not allow you to manage multiple relations between objects (e.g., a parent with several children). These are the 1-N relationships! 

Examples of Data Modeling in E-commerce 

A flat model will allow you to log the last interaction (or a little more but you will have to define this in advance and once you have, you can’t change it). This is all very well if you just want to send a customer a reminder about his last abandoned shopping cart or an email about his last purchase. 

To give you a better idea about the difference between a flat and a rich model, the best way of explaining it is the example of the bookseller who recommends Jonathan Franzen’s latest book based on your last online purchase (unfortunately you bought it as a gift for someone else) while all your previous purchases were in your preferred genre, namely spy novels. 

Examples of Data Modeling in automotive 

Another example: a company specialized in the maintenance and sale of car accessories. A customer had his car serviced one year ago but bought accessories in the meantime. If this company uses a solution based on a rich model, it can create a scenario based on the service order and send a reminder that the car needs to be serviced soon. If the company uses a flat model and only logs the most recent purchase (wiper blades), it will only be able to send the customer an email about windshield washing liquid and will not be able to trigger the customer to return and have his car serviced. 

Examples of Data Modeling in Marketing Automation

Aligning Data Modeling with your IT

Usually your own IT system already works with rich data models. Data exports from one rich data model to another are often easier and require less development, unlike a flat data model which must be fed with more complex requests, which often require aggregation, to be performed by your IT Department. Finally, every time you have a new idea about how to optimize your campaigns, you will have to ask your developers to do more work. In other words, you’re not making any friends there. 

Good to know: obviously you don’t export all the data in your IT system but only the key data you need to activate scenarios, the so-called Smart Data.

Aligning Data Modeling with your IT

Discover the infographic

We condensed the information into an infographic. Don't waste any time, discover it now.

Discover the infographic