Knowledge is continually evolving, with new designs and discoveries leading to innovation across all industries. Sellers of technical products or components have to invest in engineering and solution capabilities or be able to harvest and repurpose the expertise from elsewhere in their industry. Due to the complexity of fast moving technologies and ecosystems of suppliers and designs, customers frequently look to their suppliers to educate and guide them.
The challenge to a supplier of large numbers of complex products is to design information structures (knowledge architectures, knowledge processes and the supporting information architecture) that can be extended and quickly evolve to help buyers navigate a complex engineering landscape.
Integration of content
Designing information structures is a content integration problem, and taxonomies can be designed to support this challenge. Knowledge created across the organization needs to be harvested at the point of creation and then packaged so that it can be surfaced where it may be needed. This process must be carried out in a way that allows for serendipitous discovery of knowledge as well as the intentional navigation and retrieval of that knowledge.
Discovery can be achieved through the correct associations of concepts through use of an ontology which is designed to codify knowledge relationships. In this way, a piece of content may be tagged (either manually or through automated means), and the concept/topic relationships will cause that content to show up in appropriate locations. This application of taxonomy is more subtle and designed specifically to support the needs of a variety of users searching through large numbers of products. Product data architecture is only part of the picture.
Product knowledge and content architecture, integrated with personalized navigation and search, is the next level of user experience enabled by taxonomy.
Understanding the customer need
The best way to set up appropriate categories in a taxonomy is by having a good understanding what is important to a customer. Medical lab equipment designers don’t need to see terminology, categories or product topics related to petrochemical plant actuators but process control lab equipment designers may find those concepts important to their work. Consider these questions when designing your taxonomy of customer needs:
How are customers described in terms of the types of issues and challenges they deal with?
What types of problems are unique to their industry?
What do they have in common with others? What can be defined about their sub-field of interest or area of specialization?
Can common problems be identified for one group or another?
Personalization and contextualization at the B2B technical level is complex, challenging, and powerful.
Mapping the user lifecycle
Developing the correct taxonomies and knowledge and information architecture also depends on understanding the lifecycle of their purchase behaviors. A great deal of work may have already been done before the user comes to your site. Questions that might inspire the direction of your ontology include:
What brought them to your site?
What types of problems they are trying to solve?
Where are they in the process?
Are they at the preliminary stage with a clean sheet or have they already made decisions that will now constrain their choices?
What information do they need?
What else do they need to move forward in their design or purchase decision?
Have they already researched three other options and are now making the final determination?
Or are they a long-time customer who has a new problem with a new set of design requirements?
Each of these stages of the customer lifecycle may have different information requirements. Understanding this lifecycle can inform design decisions and help detect the subtle signals through navigation, search terminology, or site conversions in the form or downloads or calls to action. This electronic “body language” provides clues that, if properly read and interpreted, can lead to better results and new customers. Not understanding and responding to these clues means missed opportunities.
Search as recommendation engine
Search needs to be considered as an engine that interprets all this information about your user – their interests, industry, and problems as well as past and real-time behavior -- and then provides a suggestion based on a keyword or phrase trigger.
Think of search as the start of a dialog, but not the beginning of the conversation. All of the things that you have gathered about the user up until this point become inputs that add context to the dialog. The conversation has already been taking place when the search term is entered. If you have been listening.
Detecting electronic body language is similar to being at a gathering or colleagues and watching how people interact. You take into account many aspects of the situation – the nature of the event, how people are dressed, the tone of their questions, how they interacted with others – before you answer the question. You have been given contextual clues that can inform your answer. If an individual looks you in the eye, has a strong handshake, is well dressed and distinguished looking, and appears to command the respect of other senior people in the room, this conveys something different about the individual than if they were shy, quiet, eyes downcast and poorly dressed. Will you answer the question differently? Probably. The context of electronic user behavior is just as revealing as body language of people at an event. However, it has to be considered in the entirety of other signals and analysis.
The role of ontologies
The term “ontology” describes a domain of knowledge along with relationships between concepts. In the case of a list of products, categories, solutions and industries, the relationships between those concepts can be used to bring together the most appropriate combinations of products given the trigger of a keyword search. Ontologies are the source of all of the organizing principles of a modem ecommerce site, and contain the representations of knowledge specific to the expertise of the organization. An ontology is at the core of the business and its differentiation. It goes beyond taxonomies because it describes not just categories but relationships and associations among different information components. These are in many cases as important as the categories into which each product falls, whether it is the type of product or its features.
Use cases as assets
Use cases are powerful but not well understood or well leveraged in most organizations. A use case is not a single entity. Use cases and user scenarios are assets that need to be developed over time as libraries that are maintained by subject matter experts and used to test assumptions about designs and user needs. They have to be specific, and can number in the thousands over time. They are abstracted to generalized scenarios and incorporate variables to describe hundreds of variations on a single theme. Use cases become the objective test about what an organization enables for their customers and whether they are correctly serving them. They form baselines for usability testing and guide interventions for ongoing improvements.
Taxonomies and ontologies are synergistic
Taxonomies form the foundations of product organization for websites but are also the foundation for knowledge architecture, metadata schemas and search applications. In some cases, a taxonomy may not be seen by the user but be behind the scenes to improve relevance of search results. Users come to a website to solve a problem and satisfy a need, and that process requires that the correct information be provided for them so they can make their selection easily and intuitively. That requires developing not just navigational constructs but identifying factors that enter into a decision or influence a selection. Ontologies therefore become the knowledge scaffolding that contextualizes user choices and selections and narrows their focus by eliminating less relevant possibilities.
This capability requires that organizations establish the correct foundation of practices across a number of related disciplines that holistically comprise the information ecosystem of the enterprise. Getting users to the exact information they need means knowledge engineering an ecosystem that they can traverse easily and efficiently.
To learn more about the data architecture approaches that support ecommerce personalization, check out these additional resources: