Why Does In-Session Personalization Matter in Retail?
by Brian McGlynn
Online shoppers everywhere have become increasingly savvy when searching for relevant information on products and services. What happens when they enter their search query on your website? Do they receive relevant information based on their search or do they see generic information? Today’s shopper expects to see results that speak to them, which is why showing them your brand understands their needs is so important and will send them down the right path. Enter the age of search personalization.
AMcKinsey report highlighted that a majority of customers (71%) now expect companies to deliver personalized interactions. When brands don’t deliver, 76% reported frustration with the company. This is not the emotion we are hoping to foster and it doesn’t bode well for profitability – in fact, a recent Coveostudyfound that visitors bounce because of poor experience. Most (68%) consumers won’t make a purchase if they can’t find the products they’re looking for, while 51% will leave a brand’s site due to poor website navigation. A limited number of product reviews, or none at all, will lead 36% visitors to abandon the online store, while 32% of respondents said the same for a poor mobile shopping experience.
The payoffs for delivering a personalized shopping journey make the effort worthwhile. Offering personalized search delivers positive brand perception, customer engagement, and customer loyalty. And it drives revenue. The McKinsey research found that companies that excel at personalization generate 40% more revenue from those activities than slower growing companies. Hearing all this may leave you wondering how to personalize customer experiences without a treasure trove of data. Delivering personalized search for your customers doesn’t have to involve huge volumes of data or even previous customer information.
History-based vs. In-session Personalization
While the easiest path might entail access to unlimited amounts of data from repeat shopping sessions and a growing database of user information, this history-based approach isn’t always attainable. Unless you happen to have a database with detailed information related to a customer’s account, search history, prior interactions with product listings and ads, activity across other websites, social media, etc., you might be wondering – what is the best approach to delivering personalized search results with limited or no data?
Fortunately, there are a multitude of ways to personalize results by gleaning pertinent information during a customer’s visit to your website, without any prior data associated with that visitor. In-session personalization seeks to determine user intent with a smart personalization algorithm that captures every action a customer performs on the site in order to deliver relevant results. The behavior shoppers display as they move from one product to the next can indicate a need. Making the most of that information allows brands to determine that customer’s intent and therefore offer them the most relevant search result.
Brands can also look for ways to deliver in-app contextual help and surface documentation and other rich media based on a user’s specific context – i.e. the current page they’re viewing and what they have already looked at up to that point. Beyond a customer’s initial search, brands can also look for opportunities to improve the experience within the product. This could manifest itself in the form of dynamic in-experience help. This extra level of personalization contributes to a strong customer experience and therefore greater product use and adoption, especially if the brand’s search UI can be integrated into different applications in order to surface impactful content in a contextual manner.
A few technical tactics to determine customer intent online:
Time and Location - Time of day or even time of year (season) and geo-location may offer personalized opportunities to cater results to the visitor.
Query auto-completion and auto-correct - Human created searches can often contain typos, offering intuitive spelling corrections and auo-completion of search terms not only saves the visitor time and increases the likelihood of a successful purchase.
Predictive category suggestions- An AI-driven personalization system uses machine learning to make predictive category suggestions whenever appropriate. For example when a visitor browses running gear listings and then enters the term “shoes” in the search bar, the system could then suggest the “running shoes” category.
Discovery tags - A mobile-friendly approach to faceted search and navigation for smaller screens. The personalization system can try to pinpoint the user’s intent with quick filter tags to narrow down the product choices.
Dynamic facets and filters - Combining product data with natural language processing enables a search engine to locate products in the same way people think about them. There may be two words that imply the same product, i.e. pants and trousers. Or a user may place a descriptive word in their search such as a shape or size. A well organized and mapped product catalog will help the system to serve up relevant content quickly.
User clustering - Using refined similarity models, the system tries to predict user affinities based on interests that become apparent as they browse. For example a series of searches may indicate the user is a parent who is planning a camping trip and also cares about environmental factors. Capturing this data for one-time users who are not logged in is possible by assigning a temporary user ID for the session or until it expires.
Intent aware product ranking - Search results are fine-tuned to match the prior behavior of a customer either within the same session or a previous visit. For example, two visitors from the same demographic who each search for “shoes” may see different category or product results based on their prior interactions.
How AI can support in-session personalization
As new advancements emerge in the realm of Artificial Intelligence (AI), brands continually gain new ways of maximizing in-session personalization. For example, generative AI applications, such as OpenAI’s ChatGPT, have recently received much interest and buzz – and there’s already a use case for retailers with basic question answering in customer service scenarios; yet there are broader ways that AI can support in-session personalization as well.
AI advancements improve upon personalization capabilities with crawling modules and custom connectors, ranking rules management, dynamic facets, and headless UI components. As AI continues to evolve it will become increasingly impactful for in-session personalization.
Another example is using advanced applications of AI for a recommendations system that’s based on product associations and driven by in-session actions. This type of system should be able to anticipate what an unknown visitor needs and also provide it in the moment based on the initial query and related details. The key is to share recommendations for similar products that others are buying, but more importantly to base guidance on the shopper’s own in-session actions and the product information associated with those actions.
How do you know if it’s working?
How should brands measure the success of their personalization efforts? A first step is to determine what the KPIs will be. The goal here should ultimately be to switch the company mindset to that of one that is helping people buy versus placing an emphasis on selling products. This is where looking at statistics like Revenue Per Visit (RPV) vs Conversion Rate (CVR) can tell a better picture. For example, it’s easy to increase Conversion rate - have a sale. RPV is ultimately what will move the needle. What is driving and motivating efforts will also determine if your customer sticks around, or recommends your brand to others.
Beyond the more quantifiable benefits, it would also be wise to place an emphasis on creating programs and communications that are rooted in authenticity. Savvy marketing teams will take the time to understand the customer journey, which will, in turn, help both the customer and the brand win. Placing the customer experience at the center of your product and marketing programs will build stronger communities, customer relationships, build your brand reputation, and promote inclusion.
Gone are the days of relying on pre-defined marketing personas to drive customer experiences. Preconceived notions, although well-intentioned, leave customers feeling misunderstood. Think about the difference between white-glove service at a specialty store versus hundreds of aisles in a warehouse-like store with no help. Intent-driven, session-based personalization allows you to create relevant and personalized user experiences with minimal data so that all customers, whether they’re a first time visitor or VIP, receive a great shopping experience.