All Posts

Personalization - 3 Ways to Use Data to Guide Decisions

Personalization comes in multiple shapes and forms, many of which businesses can put to effective use. But they shouldn't make the mistake of launching all of them at once. An incremental approach works well here. And a good place to start is product hierarchies.

Why Start a Personalization Initiative With Product Hierarchies?

The way an ecommerce site organizes products is a primary mechanism for customers to discover products and solutions. When we know what we want, we tend to retrieve through search. When we don’t know what we want, we discover by browsing. People often shift from retrieval to discovery and back again multiple times during the course of their journey, and both of these mechanisms need to work in tandem on the site.

Search alone only works when people know what they want and know what terms to use to accurately and unambiguously describe what they want. The challenge is the search terms people actually use are short, ambiguous and an approximation of the searcher’s real information need. Therefore, after retrieving the search results, the user often needs to fine-tune the results using browsing and refinement via navigation. Faceted filtering appears as navigation but uses product attributes to refine or expand results (size, color, brand, price, application, solution, etc.). Search and browse use cases inform the product hierarchy and filters, which are enabled by the product data model.

See also: 5 Key Product Taxonomies and How They Drive Your Business

A Persona vs. an Audience Based Approach

The ways that people browse or traverse the product hierarchy vary depending on how they think about what they need and how they go about fulfilling that need. What is most important to them? What do they care about? For B2B customers, an engineer cares about functionality and technical specifications, while a procurement manager cares about pricing, supply chain reliability and performance guarantees. Each user’s “mental model” varies depending on background, knowledge, objective, and communication and work styles, among other personal attributes.

Target audiences fall into segments — engineers and procurement managers, in this case — but members of a given audience might have different work or communication styles. What an engineer fresh out of university cares about will be very different from the veteran engineer with 20 years of field experience. New engineers may have different expectations about how they interact with technology or how they collaborate with colleagues. In the case of consumers, some segments seek a trusted brand, while others might want the best price. Product hierarchies should allow for navigation paths that are dynamic enough to accommodate different types of users even within a specific audience segment. These different types of users are represented by personas — a description of an archetypical user and their personality with day-in-the-life scenarios as well as a narrative about their background, education, family status, personal goals and more.

An audience can be made up of multiple personas and a single persona can represent multiple audiences. So, should design and testing be based on audience or on persona? The answer depends on whether the audiences have more things in common or personas have more in common.

Let’s go back to that audience of engineers. The audience would undoubtedly include engineers who differed from each other across multiple dimensions. They would reflect different personalities, ways of working and communication styles for whom one hierarchical structure might be preferable over another. The question is whether “engineer-ness” is more impactful than their other characteristics.

Optimizing for one audience segment at the expense of others — engineers versus procurement managers, for example — is clearly not ideal. Another option is to design for the least common denominator. Develop a common structure that satisfies across multiple audiences but does not optimize for any one of them. Or serve up multiple navigational constructs for different audiences, either through self-selection of role, need, industry, application, or other organizing principle, or through authenticated user profile information that contains those attributes. This approach is more nuanced, and has to be tested to ensure that the needs of different audiences are understood and that their choices are not being incorrectly limited.

Ultimately, navigation should be personalized based on the results of testing various designs and the resulting behaviors. How those insights are activated to produce the experience depends on organizational process maturity, information architecture design, and the available technology stack.

3 Mechanisms to Use Data to Guide Personalization

1. Simulation to refine design

During the design phase of a project, different ways of organizing products are tested across scenarios, use cases, and tasks for different audiences and specific personas. The objective is to measure success or failure of task completion as well as the path that users took to achieve their objectives. Testing usually shows that different personas/audiences have very different needs, and those different needs are served by a different experience.

2. Data from onsite user behavior

An approach may work in testing but it has to be validated or confirmed after deployment. As changes and course corrections are implemented, those changes also need to be monitored. Do people click through to product detail pages from product category landing pages? Or do they resort to search because the landing page does not contain the information they need? Or do they bounce out to a competitor's site? Do they pick up the phone and call their rep to place an order?

On-site behavior changes should trigger design evaluations and suggest updates. A metrics-driven governance playbook can be developed to guide oversight and form the foundation for more routine, rules-driven updates.

3. Real-time response to user behavior

This is the holy grail of personalization. The goal of personalization at scale is to embed human knowledge and skill in the digital infrastructure. The system should read digital body language just as a skilled salesperson reads physical body language and responds appropriately. This ability requires that knowledge is captured and componentized, that customer data models are rich with characteristics represented by metadata (all of the details we can personalize on–location, preferences, role, topics of interests, industry, objective, etc.), that product data contains the attributes that different audiences require, and that there is a detailed model of the customer journey that represent moments that matter, micro-decisions, and responses to various messaging channels. It also requires an orchestration and optimization engine, which depends on machine learning, to interpret the various signals that are coming throughout that journey and to associate messaging combinations and recommended content and products in real time.

These approaches differ in the timing at which insights from data can be acted upon (or activated). For simulation, the timing can be days or weeks for user behavior, it may be hours to days (or for some large organizations, weeks); and for real-time response, it is essentially instantaneous or near real-time. Clearly, the more quickly that insights can be activated, the better.

A Foundation for the Personalization Journey

Data about behaviors is an input to design decisions and is essential for ongoing refinement of the experience. More sophisticated personalization mechanisms build on this foundation. Personalization is a process of anticipating what users need and surfacing information that meets those needs. Starting by adjusting navigational choices based on audience needs can evolve into more complex and real-time personalization of content, products and solutions that will one day emulate how human experts interact with customers — guiding them through choices and selections unique to their challenges and situation. Starting with straightforward navigational constructs is a solid, easy to validate mechanism for beginning the personalization journey.

***

A version of this article originally appeared on CMSWire.

For a look into how digital product data, managed in a Product Information Management (PIM) system, can be used to power product search, search refinement, and product comparison.  read our whitepaper: B2B - World-class Customer Experience Requires World-class Product Data.

Seth Earley
Seth Earley
Seth Earley is the Founder & CEO of Earley Information Science and the author of the award winning book The AI-Powered Enterprise: Harness the Power of Ontologies to Make Your Business Smarter, Faster, and More Profitable. An expert with 20+ years experience in Knowledge Strategy, Data and Information Architecture, Search-based Applications and Information Findability solutions. He has worked with a diverse roster of Fortune 1000 companies helping them to achieve higher levels of operating performance.

Recent Posts

Conversation with ChatGPT on Enterprise Knowledge Management

In another article, I discussed my research into ChatGPT and the interesting results that it produced depending on the order in which I entered queries. In some cases, the technology seemed to learn from a prior query, in others it did not. In many cases, the results were not factually correct.

The Future of Bots and Digital Transformation – Is ChatGPT a Game Changer?

Digital assistants are taking a larger role in digital transformations. They can improve customer service, providing more convenient and efficient ways for customers to interact with the organization. They can also free up human customer service agents by providing quick and accurate responses to customer inquiries and automating routine tasks, which reduces call center volume. They are available 24/7 and can personalize recommendations and content by taking into consideration role, preferences, interests and behaviors. All of these contribute to improved productivity and efficiency. Right now, bots are only valuable in very narrow use cases and are unable to handle complex tasks. However, the field is rapidly changing and advances in algorithms are having a very significant impact.

[February 15] Demystifying Knowledge Graphs – Applications in Discovery, Compliance and Governance

A knowledge graph is a type of data representation that utilizes a network of interconnected nodes to represent real-world entities and the relationships between them. This makes it an ideal tool for data discovery, compliance, and governance tasks, as it allows users to easily navigate and understand complex data sets. In this webinar, we will demystify knowledge graphs and explore their various applications in data discovery, compliance, and governance. We will begin by discussing the basics of knowledge graphs and how they differ from other data representation methods. Next, we will delve into specific use cases for knowledge graphs in data discovery, such as for exploring and understanding large and complex datasets or for identifying hidden patterns and relationships in data. We will also discuss how knowledge graphs can be used in compliance and governance tasks, such as for tracking changes to data over time or for auditing data to ensure compliance with regulations. Throughout the webinar, we will provide practical examples and case studies to illustrate the benefits of using knowledge graphs in these contexts. Finally, we will cover best practices for implementing and maintaining a knowledge graph, including tips for choosing the right technology and data sources, and strategies for ensuring the accuracy and reliability of the data within the graph. Overall, this webinar will provide an executive level overview of knowledge graphs and their applications in data discovery, compliance, and governance, and will equip attendees with the tools and knowledge they need to successfully implement and utilize knowledge graphs in their own organizations. *Thanks to ChatGPT for help writing this abstract.