It was Nature’s editor Clifford Lynch to be the first to mention Big Data in 2008. In 2016, Big Data was listed among the reasons for Brexit and Donald Trump’s election victory. In 2018, such heavyweights as Facebook, Google, and Amazon cannot do without Big Data.
Why did the collection of unstructured data become the major weapon in the competitive race? The AI Conference press service interviewed Denis Afanasyev about the issue.
Denis Afanasyev is a digital marketing expert. He has 15-year experience in IT and is focused on marketing tools for business promotion in a digital environment. In July 2014, Denis took over CleverDATA, the leader in the sphere of analysis, Big Data, and marketing automation.
Interviewer: AI Conference (AC).
Speaker: Denis Afanasyev (D.A.).
AC: Hello, Denis. There exists an opinion that Big Data would gradually enter all our life spheres, just like electricity once. Tell us please in which segments are Big Data technologies commonplace and which industries are to accept them?
D.А.: Finance, telecoms, media, advertising, and e-commerce are popular spheres of Big Data implementation. Related companies accumulate huge amounts of data and actively use them. Big Data allows to optimize various business processes, search for the client information, and receive recommendations for personalized communication.
However, for some industries like building and agriculture, huge amounts of data are not typical. It is mostly due to the long-lasting product development cycle of related companies. Nonetheless, the wide-spreading digitalization that we currently observe is expected to grow. So, both agriculture and building are highly promising in terms of Big Data.
AC: What affects the speed of Big Data implementation in various industries?
D.А.: In most cases, the speed of new technologies implementation depends on the expected results. Other aspects include the sphere of company’s activity and the complexity of business processes that require optimization as well as its ability to collect and store data.
That is why in many large companies it is innovation architects who are responsible for the inclusion of cutting-edge technologies in business processes.
It should be noted that work with data is not a one-day project but continuous development. New data resources, various machine learning models for more precise recommendations emerge on the market every day.
AC: You outline certain tendencies on the data audit market. Why are approaches toward data collection changing?
D.А.: The amount of data increases, and so do the cases of its implementation in business. The appearance of new sources, dynamic integration of online and offline data allow to analyze consumers in detail and get a broad picture of their buying habits.
Meanwhile, news feeds sometimes have us on the edge of our seats by announcing another data leakage, new vulnerabilities in particular services. Regulation of such issues requires improvement of the current legislative framework.
For example, the recently effective General Data Protection Regulation (GDPR) is applicable not only to the European market but to any company that can collect and process the data of EU citizens.
What is more, the regulation gives a maximally broad definition of the notion ‘personal data’, including ‘cookie’ and ‘MAC address’. This leads to substantial changes in the way the heavyweights work with data.
Big Data isn’t clearly defined in Russia yet. The same situation is observed with the special regulation of data collection, processing, usage, and trading. Meanwhile, companies representing different industries establish partnership in terms of data exchange.
Integration of different sources that provide data about consumers opens new opportunities for work with a customer base, improves targeting, personalizes communications and as a result optimizes the cost of client attraction and retention.
AC: Your company CleverDATA is now focused on the automation of marketing communications and online advertising. Tell us please what client data do intelligent marketing systems collect?
D.А.: Primarily, personal data is managed by a company. It contains the information about website and mobile app hits, CRM consumer behavior, loyalty programs, etc.
The data about the audience interaction within marketing communications is important to data analysis. The following criteria would help better understand client interests and establish a single sequential strategy of internal communications: how clients opened an email, the number of hits, browsing habits, how many times they saw a banner ad, what link they followed and which one ignored.
Extra info can be found on external sources. This function is available on various aggregators and a data exchange that interconnects many Big Data suppliers.
AC: How do you use this data for personal information personalization?
D.А.: When data from all available sources is gathered and processed, you need to create a detailed form of each consumer or a website visitor. It should include a profile with social and demographic parameters, sports activities, music and film preferences, hobbies, children, country estate, a car, intentions to buy some product or service, etc.
New knowledge about the audience helps generate a lot of microsegments and launch targeted advertising campaigns with offers consumers would find interesting.
Besides, by dividing the audience into several segments (for example, sex and age), it is possible to customize the website content. For instance, when entering the online-store of bicycles, women see women’s bicycle models while men see men’s products, children – bicycles for the youth.
AC: What are the advantages of AI-powered systems over marketers? In which spheres do human workers show a better result?
D.А.: Compared to human workers, artificial intelligence is capable of processing and checking a mountain of data. It’s difficult for your brain to understand and see hidden regularities in huge amounts of data, which is an executable task for AI.
However, AI is good at solving clearly set tasks: one algorithm searches for an optimal route while another – product recommendations. AI shows better results in some cases but of course, it’s still far from the power of a human brain.
AC: At AI Conference, you will be speaking about the role of machine learning in marketing communications. In brief, what aspects will you feature?
D.А.: I will dwell on the role of machine learning in establishing marketing communications with maximum personalization and a minimum part of a human in this process. I will provide my own case studies demonstrating how customer data may help recognize client preferences, the best time, right frequency, and discounts.
I will also compare the results of the work of several machine learning models, including neural networks for product recommendations.
To meet Denis Afanasyev and other speakers, register to AI Conference scheduled for November 22 in Moscow.