Five ways retailers can compete with Amazon’s record breaking online sale

Amazon’s sales holiday season on July 12 achieved record sales for the company in only its second year. It is estimated that U.S. sales alone were as much as 50% higher than the year before, while global sales were 60% higher. From selling 95,000 TVs to small businesses recording 3X the number of sales than during traditional holiday periods, the event caused many retailers to rethink their traditional sales tactics.

Trying to match Amazon’s approach, however, will fail because general retailers would find it challenging to compete with Amazon.Retailers need to think differently and blend analytics with behavioral science to lure customers in this era of high competition, shift towards digital, and changing customer behavior.

Why do people buy so much from Amazon? Aside from price, the science of influence theory and analytics is vitally important. Shown below are the overall strategic principles that guide most retailers along with observations regarding what they should be doing:

    1. Feeling left out:In the modern retail climate, people are conditioned to buy only when things are on sale. In the   average mall today, something is almost always on sale, which actually means that nothing is ever on sale. According to Manmit Shrimali, founder of Dextro Analytics, “Customer behavior is very irrational. Having regular sales not only hurts the bottom line but also destroys the basics of consumer behavior—urge of urgency and relevane.” Retailers need to stage their sales so that customers will feel they will lose out if they don’t take action. Feeling left out is often cited to explain why customers do what they do. The artificial intelligence–based algorithms can be used based on historical evidence related to the sales season, shopping behavior across categories, and SKU performance. Such knowledge can then be used to devise a sales system to help marketers decide when to have a sale and then create a very targeted sales campaign that makes consumers want to shop.

    2. Making the customer feel special:Customers used to be lured through loyalty cards, but virtually every retailer today offers a loyalty card. Retailers are very familiar with shopping behavior across sales channels. Such information can be used to better understand what is needed to make their customers feel special. As Plato said, “Human behavior flows from three main sources: desire, emotion, and knowledge.” Retailers need to give love to get love. The data that retailers have can be blended to understand what makes their customers feel privileged, which allows them to develop a successful selling strategy.

    3. Using differentiating analytics to optimize business functions:Retailers can always keep lowering their price, but such an approach is like winning the battle but losing the war. While price is certainly important, retailers can use analytics to improve their business functions that will affect pricing and eventually profitability. For example,
      • Optimize the supply chain-from stocking the right item to choosing the best route for delivery, newer algorithms can significantly outperform traditional ones. Different models can provide very different insights on the same data. Retailers can differentiate their supply chain through deep algorithms that can lead to measurable cost savings.

      • Predict what items may be returned:Returned items create a major problem for retailers, but they have historical information from past holiday seasons regarding shopping behavior, for example, who returns what and when. As noted by Ajith Govind, founder of Dextro Analytics, “Through Markov-based time series analytics, we have been able to help retailers predict with high precision which SKU is most likely to get returned. With such precision, retailers can reduce the returns by as much as 18%.” Retailers can leverage such precision to put on sale only those items that will likely not be returned. Determining what these items should be, though, can be challenging. It requires complex feature extraction from a massive number of SKUs over several years, but today’s computing power and deep learning algorithms make this possible.

      • Dynamic segmentation: By moving from a static to a dynamic segment through artificial intelligence dynamic-activity-segmentation, retailers can now personalize the product, price, and promotion not only at the segment level but also at a micro segment level. We are living in an age where customers no longer fit into a single segment. We need to use dynamic micro segmentation that can generate insights much more granularly and at a deeper level.

    4. Improve predictions on what to show to customers:When customers land on your website, you have only one opportunity to show the most relevant information at the inception and when they search for products. If you don’t get it right with high accuracy, you will not only lose the sale but also generate detractors of your brand for wasting their time. Being able to recommend the right product is Amazon’s biggest asset. At Dextro Analytics, we benchmark retailers on their accuracy to show the right product. Through our first two-quarter analysis, we conclude that there is significant variation and opportunity to improve the personalization of e-commerce or online retailing. Again from Manmit Shrimali, “Learning from millions of clicks across SKUs over time periods provides a tremendous opportunity to understand behaviors, but such learning requires deep expertise and an algorithm to untangle the complex relationships.”

    5. Going beyond big data: A famous presentation made by Peter Norvig, Google’s director of research, carried the following title: “The unreasonable effectiveness of data.” Retailers can access a vast amount of data today through big data technologies. They can connect millions of data points to find the patterns, but they also need the right data. Novel technologies are available that can blend data, and these outperform traditional technologies to process that data and mine emotional and behavior intelligence. For example, they can do the following:

      • Pricing an SKU in odd figures tends to yield 6% more sales than pricing the product in even numbers.

      • Giving reference points helps to sell the same product at a higher price.

      • Offering more variety triggers complex decision-making that the human brain often tends to avoid.

As offline sales continue to decline and consumers become more and more irrational and price conscious, retailers have a tremendous opportunity to use data and physical location to shift to advantageous online selling.

To find out how you can compete in a digital world, please contact :
Prince De
Director, Client Success Dextro Analytics Inc.
+1 647 273 2309 (Canada)
+1 206 460 1800 (USA)

Leave a Reply

Your email address will not be published. Required fields are marked *