Is marketing an art or science? Less than five years ago most CEOs would have classified marketing as a creative industry, requiring intuition, imagination, and talent, just 5% would have mentioned big data as the core of their success. These statistics are rapidly changing in favor of a data-driven approach, showing that over 60% of companies are using Big Data .
The questions on everyone’s lips are: Will data dictate the way for marketing or will it support it? Will creativity and human touch still be part of the advertising process, or will they be completely replaced by algorithms?
Data: between historical assessment and predicting the future
Until this decade, data was used more for evaluation purposes. Typical tasks included looking at how well a campaign performed, what the costs were and returns on investment. Sometimes data was extrapolated for predictions. Currently, there is a paradigm shift, and data is considered a competitive advantage that gives customer insights and is becoming the guiding principle behind decisions.
Artificial intelligence or even simple automation are helping companies save millions of dollars and expand to new markets, engulfing new customer segments. Companies do not wait anymore for the end of the quarter to look at data but have incorporated in real-time and even predictive responses in their marketing strategies.
Benefits of using data driven marketing
Sentiment analysis in the voice of the customer
Before the explosion of social media, marketers relied on the results of surveys and focus groups as a proxy for what customers wanted. These results were biased most of the time by the small sample of potential clients available and the researcher’s methods. The true voice of the customer was usually heard only when it was too late, as a complaint.
Data-driven machine learning algorithms  can evaluate a review, a comment or a question in a forum and determine the topic, the product it refers to and the sentiments of the customer. This is almost like conducting an interview for each client willing to talk to you. The reliability of such tools is high since these are all trained by using real user-generated content. The best part is that the algorithm can pick even the feelings expressed by emoticons, as this method of communication is replacing facial expressions online.
A/B testing and Bandit testing
Deciding which picture or tagline to use is an ongoing dilemma. The A/B testing method of selecting the best course of action has been used in medicine for decades under the name “controlled testing.” Generically called A/B, suggesting a choice between two alternatives, it can be generalized to A/B/n to accommodate as many options as you need. Specifically, you create two or more variants of an ad and run it on a platform of your choice on a sample group to evaluate the best performing variation.
From a marketing perspective, gathering this type of data helps save money that would be wasted on an underperforming set of ads. The real power of data is leveraged by bandit testing , an automated generalization of the A/B method. Instead of splitting the sample evenly, the computer redirects visitors dynamically to the preferred option of the moment, keeping the other in the background with lower visitor influxes.
Multiple channel strategy
An excellent process is a stable and predictable one. That is just what data promises to bring into marketing. The ability to create scalable, repeatable processes and to estimate responses and financial results. The great thing about data is that the lesson learned in one channel can be successfully repurposed for another one for a multichannel strategy .
For example, if Google Analytics shows that the visitors of the website are within certain demographic determinants, this is a good estimate for targeting similar users via Facebook ads, or, on the contrary, it gives insight about a category, insufficiently reached.
Fast forward through the sales funnel
Building a strong client relationship is easier with data. By learning quickly what their preferences are, predicting behaviors and always having the right answer even before they ask – you have a solid and fast way of turning prospects into customers and customers into evangelists.
Data helps create customer clusters, anticipate trends and be ready to serve faster and better. Furthermore, such solutions can be fully automated by using customer intelligence  solutions which take information from different existing sources like census bureaus, sales points, CCTV, social media and aggregates it into a single relevant dashboard, ready to answer business questions.
Challenges of data driven marketing
An ocean of data
Taking the best decisions requires crystal clear focus and the 4Vs of Big Data (volume, velocity, variety, and veracity) are not helpful. A company which aims to have a competitive advantage from data and uses it to become more customer-centric needs to define a strategy, a handful of relevant KPIs and the necessary raw data. The one-step-at-a-time approach and small changes can yield more results than a holistic Big Data transformation. To leverage the full potential of this method the company requires better data housing methods, real-time analysis, and full support, therefore, it is best to seek advice before launching a data-driven project.
Losing the human touch
The biggest fear of specialists, not only marketers, is related to the uniformization of techniques that may kill the human touch, creativity and ultimately differentiation. This scenario looks counter-productive in the long run. However, this is a highly unlikely situation, since big data systems use input from real users which are so diverse it is impossible to end up with a uniform result.
At this moment, marketing built on data is far away from turning everything gray. In fact, it is emerging as the best way to make the client feel important and well-catered to. This approach to marketing is creating micro-niches, where every customer is valued, and a customized solution is available for them regardless of the problem. Furthermore, that solution sometimes just pops up on your screen as if someone read your mind. They haven’t read your mind, they just analyzed your browser’s cookies.
About the Author
Jasmine Morgan is a solution architect with 8+ years in software consulting. Since 2013, her advanced focus is IT solutions for marketing and financial sector.