Children dreaming of becoming astronauts... is that age over?
As it happens, Data Scientist has been defined as the “sexiest profession of the 21st century".
To deal with the complexity of Big Data, you need someone who knows how it works. It requires not only an intelligent choice of data to work with, but also the ability to process it to meet specific business needs and objectives.
At the same time, content can be receptors of data about users: you can get a lot about their interests from the way they interact with content. Someone said “Tell me what you read and I'll tell you who you are”. That is the very definition of Content Intelligence: Artificial Intelligence applied to content can give us very useful metrics on how it is performing; Intelligent DAMs are technological tools that natively complement it.
Let's discuss how Data Science will end up intersecting with Content Marketing with Alessandro Giaume, author of "Data Scientist. Tra competitività e innovazione".
Q: In your book you identified the three main skills that this professional should bring to the table. Briefly explain us what they are and how they can be applied to companies according to their data-drivenness.
A: The Data Scientist is a pretty young professional role, having appeared 6 or 7 years ago, following a famous article written by Thomas Davenport on Business Harvard Review. Little is known yet and often the skills that characterize this role are confused with others, which instead would define different professions.
Trying to clarify, the main skills a Data Scientist must have among its assets are: business, mathematics and statistics, programming.
Let's start with the last. The manipulation of text files, the understanding of vector operations, the thinking for algorithms: some key elements begin to emerge and do not seem to lead us back to a pure computer science background. Here it is also important to know how to use social engineering, to put it in words with Kevin D. Mitnik, aka "Condor", the most famous hacker of all times.
To be able to act on data, these must have been prepared. The second required competence pertains to the proper methods of mathematical and statistical sciences, thanks to which extracting meaning from data becomes possible. But a science is such when it contributes to the growth of knowledge in a broader sense. This is the reason for the third competence, that is substantial knowledge, that is business rules.
On the other site, companies are not all at the same level. Some are more data driven, some less. And Data Scientists don’t all possess the same maturity and experience. The more data driven a company is, the more it needs a Data Scientist with a full set of skills and competencies. The less data driven it is, the more basic competencies it will need.
For sure everyone needs stats and maths skills, the first to be built up. While business knowledge is particularly required by structured and used to data companies. In other words, companies that are starting to consider data as a resource, can glean business approach directly from business people.
Also hacking competence can be adopted only by those companies with a solid data strategy. Otherwise the risk is to jeopardise running business while trying to accelerate…
Q: What are, in your opinion, the advantages of applying Data Science to Content Marketing? Do you think the support of AI is important in measuring content performance?
A: Content marketing involves the creation and sharing of editorial content in order to acquire customers. The goal is therefore to create interest for a given product or service and based on this interest engage and entertain their audience, with the ultimate goal of monetisation.
The producer of these contents must forcefully be a Subject Matter Expert, able to express the strength of that product or service in the context of reference. And they must also be able to grasp the nuances that represent the real key to engage the public, users, customers.
Let’s focus now on the production of content suitable for digital use. One of our primary objectives could be to precisely identify a topic of interest to a large audience.
To do this we should probably be able to analyse large volumes of data, composed of contents of various kinds, highly variable and with a high velocity and… ops, these are Big Data. And when data are big, we need Data Science to extract value from them at the best.
This translates into the ability to access countless sources, analysing them in real-time or near real-time, deducing the themes of greatest interest in “that” moment and then enabling brands to be mainly producers of content that engages and “loyalises” customers, accompanying them with increasing frequency to purchase!
That said, knowing what customers, potential or not, appreciate, how can we control they’re not changing their mind, making our efforts useless?
Artificial Intelligence is a serious answer. Algorithms can definitely help us in following these changes, understanding from time to time which aspects of content are rewarded or, rather, scarcely appreciated.
And consequently, suggest which ones have to be produced, when not automatically done.
Q: Most of Analytics approaches adopted by marketers are descriptive, but more and more platforms now support predictive modelling. In your opinion, to what extent can the automatic personalization of web content affect business results?
A: Automatic web content personalisation is a clear direction to be followed.
We are talking about the near future of the content production strategy, populated by Machine Learning algorithms, which collaborate with human writers, anticipating any possible questions and delivering the answer to a selected audience.
Algorithms and data can provide precise information on editorial operations to be done, supporting the choice of contents schedule and specificity, identifying which of these can increase traffic to institutional sites, as well as what are the best ways of use that should characterize them.