Product and promotion recommendations

powered by Digital Sherlock

Elementary my dear Watson

How does Digital Sherlock Holmes deduce things?

Yes, the world’s only consulting detective is a walking, talking big data analytics center. His data center, his brain where everything gets stored, analyzed. The big data powered product and promotion engines are your personal version of Sherlock Holmes. He observes your customers, gathers data about them and helps solve your cases of “what does Harry like?” and “what promotions will bring back Sally to the store?”

243 Types of Tobacco Ash: Insights into a diverse customer base

Well, just try to summarize the diversity of customer base the way Sherlock does. He actually knows every single thing that is important to know about those tobacco ashes. So should you about your customers. What clicks with Sally might not click with Clyde. Following closely their purchasing patterns will disclose their unique identities. Instead of thinking all people in their late 20’s in a particular area are looking for cheese, bread and eggs to survive, big data helps you figure out each individual’s life. Maybe someone is going through a break up and someone is getting married (social media updates), someone is going on a diet and someone just wants to indulge (purchase patterns). As a retailer you need to emotionally bond with these individuals by understanding who needs new bone china and who needs a box of Kleenex. Targeting each of them with their own unique set of products and recommendations will help in fostering brand loyalty, optimize pricing (adjusts to trends and stocks, etc. in real time) and create personalized marketing campaigns.

Sherlock, timing!

Take the example of Target and their spooked customer. According to a report in The New York Times, Target followed the purchase pattern of a high schooler and figured out that she was pregnant. Even before the girl’s family! They made full use of their big data and sent targeted product recommendations for a soon to be mommy. When the father found out however it was not so much of a pleasant surprise. Leave it to Sherlock to reveal embarrassing secrets! Well then, Sherlock really didn’t have much of an understanding of social intricacies anyway. That is why he needs his Dr. Watson to remind him when he is being “a bit not good”.

According to a 2017 report by JDA and PwC 86% of retailers were planning on investing in big data.

The use of social media and big data is highly valuable in giving retailers deep insights into rich sources of customer information allowing them to create credible customer segments, while gaining insight into shopper preferences.”

Big data with beacons, Wi-Fi and Bluetooth: Conductors of light!

With big data in your hands, you don’t just see, you “observe”. You observe and know the in store browsing patterns of each of your customers. Beacons are your conductors of light in store. With their help you can figure out on which aisles your customers are spending more time and which ones they don’t visit much. These incredible little spies get you all the data you need to fill you stores with the most selling products and save money on least selling products. This knowledge helps you optimize your inventory.

Retailers are also employing heat maps to follow around their customers in the store. This helps you track movements of customer segments, like where are kids spending more time? Is it around the new DVD corner? Then the visual board should carry promotions like “10 off on 3 DVD purchases for the next one hour!” Beacons and heat maps help you understand your footfall and rearrange the shop floor planning accordingly, to get the best out of every customer visit. It also increases the chances of cross selling. Like keeping the fruits segment near the breakfast cereals and milk will encourage customers to pick them up more often.

Using geo fencing and beacons, targeted customers can be sent real time redeemable offers when they are in the vicinity of a store or a particular aisle. Sending a text saying “get 40% off on your favorite brand of jeans at our nearest store” or “one time offer! Get one box of flavored serial free with a box of your usual brand of cereal” will compel customers to check into your store.

Big data for store associates: My Dear Watson!

Helping Sherlock on his adventures is his trusted companion Dr. John Watson. For a retailer it’s the store associates in brick and mortar.

Collecting information from beacons, browsing history and purchase patterns big data Sherlock can create a 360 degree profile of individual purchaser. With the help of IoT devices (mPOS, tablets and smart phones) the store associates study and understand the profile and take action based on that. Rupert buys one liter of olive oil every month. This month he has purchased all his usual but no olive oil. At the checkout counter or mPOS the store associate asks him whether he wants his usual supply of olive oil. Rupert had forgotten to put it on his list, he is thankful that the store associate remembers it.

What your store associates do for you in store, recommendation engines do online. Take the example of Amazon. You look at a dress online, it immediately suggests other similar dresses or clothing from the same retailers as “Those who bought this also bought that”. When you add the dress to the cart, it shows other variants of the dress, shows a discount on fashion jewelry and shoes. These real time recommendations make the most of online showrooming and increase the chances of upsell and cross sale. The customer might buy the dress that is more costly but the color is her favorite, she may also add some jewelry to her cart lured by the offers.

On her next visit Amazon remembers her choices and presents more options of the same kind (new variety of jewelry, latest designs by the brand of dress she got last time).

Big data analytics: The process of elimination

When you eliminate the impossible, whatever remains, however improbable is the truth.

“Without big data analytics, companies are blind and deaf, wandering out onto the web like deer on a freeway.” (Geoffrey Moore, author and consultant)

Elimination is a crucial step in accurate deduction process. Let the predictive analytics do the math of “if A buys B, whether he /she will be interested in C or D”. Running hundreds of scenarios about why a person picks out a box of a particular serial and not the other will tell you if it is low fat? High on fiber? If it’s high fiber then will they be interested in fiber enriched biscuits too or will they be more inclined to organic vegetables. What are the more accurate scenarios that they will buy another product? Running predictive algorithms, predictive analytics find the best results which support the business in finding the right solutions to keep them customer focused.

Forecasting trends and purchase patterns by analyzing market trends helps retailers put the right products in front of the customer at the right time. Putting the right products on the shelf, online and on store windows ensures quicker sales and increased customer loyalty.

Customers are more vocal about their likes and dislikes nowadays with the advent of social media. To understand and get insights from the emotions expressed on those platforms retailers equip themselves with sentiment analysis. Preferences, frustrations, underlying dissatisfaction all can be gathered from a simple tweet or Facebook post. Sentiment analytics uses natural voice processing to derive meaning from human language.

Big data & analytics: The problem solving team

As Andrew McAfee said, “The world is one big data problem.”

From talking Barbie dolls to beacons, retailers are gaining more and more insights into customers’ lives to enhance customer centric retailing. With core analytics engines running dynamic algorithms to create real time promotions, that day doesn’t seem far when retailers would forecast what the customer is going to buy the moment she/he steps into the store. With the help of Watsons, process of elimination and conductor of lights, Sherlock will keep solving the challenges of big data.

And one day when Mrs. Hudson enters her house with hands full of groceries, Sherlock would call out, “I knew precisely what you were going to buy even before you bought it.”

“That is elementary for you, my dear Sherlock.”

Author : Shreyasee Ghosh, Research Analyst

Practice Head: Raj Bala, Principal Big Data Architect