Tech Guest Viewpoint: Personalizing Customer Interactions

By John M. Pierre, Linguastat 
 
The single greatest challenge that retailers face today is to have an individualized “conversation” with each customer and to do so on a global scale. Consumers expect a similar personalized experience regardless of channel, from the marketing offers they receive to the products that appear in search results.
 
But actually achieving mass personalization requires a number of key strategic decisions, as well as nuts-and-bolts technology tools to turn the goal into a reality. Each retailer’s route to mass personalization will differ from product type to product type, channel to channel, and even customer to customer.
 
Retailers’ first step is to determine what level of personalization is appropriate for their product offerings, as well as their customer base. Nor is this a simple one-time-only decision, but rather the basis for many more decisions For example, some customers will welcome personalized treatment when shopping online or via their mobile devices, but will prefer self-guided anonymity when they visit a brick-and-mortar store. Others will appreciate being recognized by store associates, particularly if they can help them find what they’re seeking, but will find the same suggestions intrusive, annoying or, ironically, impersonal when they are offered online.
 
Despite these complexities, e-commerce retailers have already made some significant strides in this area. According to Aberdeen’s E-Commerce Supply Chain report, online retailers have identified two key strategic actions to personalize the e-commerce experience:
 
● Increase available product data for customer review (30% of respondents);
● Coordinate product placement with customer behavior (24%).
 
Providing more information gives consumers the control they expect. Better data allows them to compare product attributes, while also enabling the retailer to up-sell and cross-sell using recommendations that stem from the customer’s search patterns.
 
Sounds good. But just how do retailers successfully implement these levels of personalization – particularly as the volume and variety of products they sell online continue to increase? One level is via product assortment. Retailers can shape the inventory they offer to various customer groups based on fixed and historical factors, such as customer demographics and previous purchases, as well as more contextual clues (i.e. the device/platform the customer is using, the time of day, and the shopper’s digital “route” to the retailer’s site.)
 
Another critical though often overlooked level of personalization involves the data and descriptive/promotional content connected to each individual product. These descriptors, which are often the last “ad” a customer sees before making a buying decision, can and should be tweaked and targeted for different audiences, pointing out the product’s relevance to the customer’s needs. When used effectively, such product-specific, customer-specific personalization can improve conversion rates while minimizing bounce and shopping cart abandonment.
 
Achieving this deeper level of personalization on the mass scale that retailers require is beyond the capacity of even the largest and most efficient marketing organizations--unless they have access to sophisticated, retail-specific content creation engines. New advances in Artificial Intelligence and Natural Language Generation are making it possible for retailers to create unique, non-duplicated descriptor variants, even for the same product. These descriptors are designed to boost online search relevance, but even more important, they are continuously refined and targeted based on dynamic metrics such as search queries, site traffic, and conversion rates.
 
So remember that true personalization is about far more than putting “Dear [FIRST NAME]” at the head of your e-mails. It’s about using dynamic content to link your products to each customer’s here-and-now needs, desires and aspirations.
 
John M. Pierre is CEO and co-founder of Linguastat.
Login or Register to post a comment.