Predictive Analytics

What are the common challenges faced by a Predictive Marketer?

Predictive Marketer
0

 

By the concept of opportunity cost, any rational person is every day faced with multiple choices, and when he chooses one over the other, it is because the benefit obtained from that choice is greater than the opportunities lost from not choosing the others. So how does this apply here? In this era of cut-throat competition and thinning margins, the need of the hour is ‘intelligence’ that would win new customers, satisfy them, maintain customer confidence, and turn them into brand advocates. In marketing lingo, this translates to Reach, Engagement, Activation, and Nurturing.

predictive marketing

But how do we achieve this? The answer lies in the complexity of the situation, this cannot be solved by simply using traditional reporting and descriptive statistical analysis, it requires greater expertise, the expertise of predictive techniques that look deeper than what is superficial and what is beyond the past and the present. So, why would a client still choose reporting and analysis over predictive methods? Why would the client prefer lower returns and not en-cash on greater opportunities? This brings us to the challenges that a person would face when he tries to sell predictive techniques in the market.

The common challenges faced by the Predictive Marketer

To understand the challenges, we first need to understand the common perception about “Predictive Analytics” in the online market.

I bet they would mostly be around ‘Geeky’, some would say ‘Expensive’, ‘I don’t need that’, ‘Sorry, I don’t have THAT much TIME’, while some would ‘Swear’ by it. These very perceptions about Predictive Analytics cause major challenges for the marketer of the service.

Let’s address four such issues.

Predictive Analytics is cost intensive but ensures better ROIs

Deploying predictive analytics requires significant investment in terms of human capital, with a strong background and experience in using statistics and a platform where the models can be built like SAS, SPSS, etc. Further, if there are multiple sources of data with their sizes exploding, driving the client to some Big Data solution with platforms like HADOOP–it might not be easy for the client to make such large investments without some guarantee that there would be a sizable ROI. But just like large battles cannot be won with sticks and stones; to make a profound business impact you need the right combination of brain, technology, time and money, and the returns would be fruitful.

Predictive Analytics requires patience

Certain predictive methods are time intensive, they require patience during execution. There are factors like statistical significance and sample sizes to meet the proper level of confidence on the insights delivered. A simple example would be,for instance, an MVT testing, which is conducted on 6 different locations on the same page and each location has 2 versions hence the required number of combinations to be tested out would be 2^6= 64 combinations and to declare a winner with at least 95% confidence and a high volume of traffic would still take sometime. If the client is not patient enough and prefers to go by his gut feeling the chances of error are broader than making a calculated guess using predictive techniques. Here is a Warren Buffett quote for people with less patience: “No matter how great the talent or efforts, some things just take time. You can’t produce a baby in one month by getting nine women pregnant”.

Predictive Analytics is not a word from God’s mouth

“There are three kinds of lies: lies, damned lies, and statistics”- Mark Twain

There are certain perceptions that float in the minds of people. They seem to consider statistics either as a tautological truth or utter lies. The problem here lies in the understanding of the subject. Statistics is the study of historical data and using them for predictions with certain assumptions and the major consideration of ‘confidence’ like saying a 95% confidence would mean that the hypothesis would hold true in 95% of the cases given that the pattern continues to hold true or the anomalies/shocks are predetermined.

Predictive Analytics requires establishing credibility

“Rome wasn’t built in a day” – Li Proverbs as Vilain

There is always a level of confidence that needs to be established before you can mature into a credible predictive analytics partner and this confidence does not come in a day. Clients would not be comfortable to make the large investments unless you prove your worth from past experience and hands-on work. But someone has to give that edge.Sometimes there are clients who have long been partners of the reporting or analytics division. The best way is to take that data, create some real startling predictive insights, and present just a section of it to create the excitement which leaves the client asking for more. With successful implementation, ask for client recommendations which would strengthen the pitch for the predictive services to new clients, serving as a testimonial. Also, provide sufficient case studies and white papers as to how predictive services can impact business across various verticals. Lastly, be confident, transparent and honest when interacting with the client and win their confidence.

A glimpse of the predictive analytics process flows at Nabler presented to the client

Predictive Marketer

Thus, the predictive marketer needs to handle the client smartly keeping the above salient points in mind while pitching. There will be clients who believe in the ancient Chinese philosopher Lao Tzu who said, “Those who have knowledge don’t predict. Those who predict don’t have knowledge.”For the rest, we will predict.

You may also like
Relevance of Predictive Analytics in the Digital World
How much data is enough data for Predictive Analytics?

Leave Your Comment

Your Comment*

Your Name*
Your Webpage