Principal Data Scientist, Practice Lead for Data Science, AI, Big Data Technologies at Teradata. O’Reilly author on distributed computing and machine learning. ​

Natalino leads the definition, design and implementation of data-driven financial and telecom applications. He has previously served as Enterprise Data Architect at ING in the Netherlands, focusing on fraud prevention/detection, SoC, cybersecurity, customer experience, and core banking processes.

​Prior to that, he had worked as senior researcher at Philips Research Laboratories in the Netherlands, on the topics of system-on-a-chip architectures, distributed computing and compilers. All-round Technology Manager, Product Developer, and Innovator with 15+ years track record in research, development and management of distributed architectures, scalable services and data-driven applications.

Wednesday, June 25, 2014

4 flavors of Data Driven Marketing

Information drives marketing. And marketing drives sells and profit. Therefore, it is paramount to possess the right information when starting and running a new marketing campaign. However marketing campaign come is all sort of shape and colors. So, what are the fundamental marketing strategies out there?

In this post, I illustrate four basic approaches to marketing. Each of these strategies has its own plethora of tools which can be used to steer the campaigns.

1 - Push Marketing


Actively promote and advertise products and services

This is the most beaten path when talking about marketing. This approach drove marketing in the past century on mass media channels such as the newspaper, the radio, the television in order to promote and advertise products.

Today, with web mobile apps have fragmented the just mentioned channels into a large number of small and big networks. Push marketing requires solid budgeting, and in-depth knowledge of networks where the products is going to be promoted. Full blown marketing campaigns are becoming extremely costly to execute, if not properly sized. One way to mitigate the cost is to define segments. In this way, the campaign can focus on specific objectives and targets.

Marketing campaigns segments can be anythings from date time filter e.g. evening hours or geo-location base, for instance, only european countries . They can be technology driven, e.g. a campaign only for android tablets, or focus on specif demographics such as young adults, car owners, etc.

What big data can do for you:
Big Data analytics allows you to reduce the campaign costs and to address only those segments, networks and channels where you want to drive growth and presence.

Marketing Campaigns Mass Marketing Multi-channel Advertising Promotions Marketing Networks


2 - Pull Marketing


Define inbound strategies, let the user reach out for the desired product.

Pull marketing, also known as permission marketing, requires to setup a network of inbound marketing paths. They can be physical like a meetup group discussing a specific topic related to the product, or digital presence such as blogs, reviews, or tutorials.

Pull marketing is normally strongly linked to content. The right content will increase the popularity of the product and boost the . Viral campaigns are a good example about how pull marketing can generate large volume of leads.

What big data can do for you:
Big Data will provide insight on the popularity and the engagement of a given pull marketing campaign. By looking at the statistics you can identify kpi's such as popularity indexes and conversely it will provide hints to better tailor the content to the given target group.


Inbound Campaign User Communities Webinars Tutorials Talks User Reviews Freemium Champions Blogs Viral Campaigns


3 - Personal Marketing



Propose relevant products for each user right when he/she needs it.

Personal marketing allows to tailor the marketing campaign to a specific profile. It can incorporate information coming from multiple angles, for instance the user mobile location, but also the demographics info for a specific user, the user preferences, the social network interaction.

What big data can do for you:
This information can be combined to build very sophisticated customer profiling techniques. By using real-time analytics it's possible to propose and recommend products right when the user needs them most.

Data science and machine learning can be deployed to tailor the offering at an individual level. These data transformations can identify the user's context and react appropriately to the user’s journey as it happens.


Context Driven Marketing Recommender Systems Trending Topics Trending Products Customer Profiling



4 - A/B Marketing


Propose theme variations for the same product campaign, in order to better appeal to a given user group.


Dive into behavioral psychology, cognitive science. Even when dealing with a single homogeneous group, such as in cohort campaigns, there is always room to explore and learn more about the user preferences and inclinations.

This sort of systems  make use of feedback loops. They oscillate between exploitation and exploration. On one side, the system tends to maximize the yield on a given metric, on the other side it will explore the space of options to find better mixes of variations and offering schemes for a given product and user group.

What big data can do for you:
Provide a set of ranked options, which address different profiles, and explore the yield by continuously adjusting the mix of proposed adverts.


Multivariate Marketing Offer Ranking Personality Match Marketing Space Exploration Behavioral Marketing.