From Relevance Laggard to Leader Becoming more relevant to your customers, communities and staff WWW.COVEO.COM 1
JANUARY 23, 2017 The Coveo Relevance Maturity Model Cheap Search is Expensive. Your customers have been taught through 18 years of exposure to Google, to expect highly relevant content when they search for it. Anything short of this negatively affects conversions, consideration, and commerce. Settling for a basic, generic search solution on your external or internal facing web sites and services directly affects your company s ability to engage with your customers or upskill your employees. 2
The Age of Relevance According to Forrester Research, we are all living in the Age of the Customer 1 - it is here and now. Customers are more mobile. They give more credence to information from peers than from companies. And they are spending more time online than ever. At the end of the day, we are all people and our consumer behavior has influenced the way we interact with brands and technology. Organizations need to recognize that this shift in expectation is just as prevalent inside their organization amongst their employees as outside with their customers. Companies need to become both customer-obsessed, and employee-obsessed, making sure that everyone who interacts with the brand has the right information, every time, at the moment they need it. In short, they need to become more relevant. Companies are beginning to recognize that relevance is the currency of our times. If you are not relevant to customers and remain so they will quickly find those of your competitors that are. Similarly, if your organization cannot provide the most relevant information to your staff, they will never be truly proficient, or satisfied and they will leave. 1 Winning In The Age Of The Customer, Gazala, M. E. with Bernoff, J., Condon, C., McNabb, K., Ryckewaert, E., Trafton, R., Forrester Research, April 6, 2015 3
The Need for a Relevance Maturity Model Organizations in the Google era know their customers and staff have inherent familiarity with finding the most relevant information through Google. And most acknowledge they are falling further and further behind their customers expectations in delivering the most relevant information, the moment it s needed. More will admit that they ve never truly provided the engaged experience their digital-native employees desire. A relevance maturity model can be seen as a set of structured levels that describe how well the systems, people and processes of an organization can meet and beat customer and employee expectations. Maturity models are used as comparison benchmarks, to help organizations recognize where they are, identify where they want to go, and prescribe what to do to get there. The Coveo Relevance Maturity Model (CRMM ) provides a continuum over which relevance maturity can be developed, moving incrementally from one level to the next. It is typically not viable to skip levels, though organizations can indeed move up multiple levels in one step. 4
Levels of the Coveo Relevance Maturity Model Organizations will commonly progress linearly through each level, at greater or lesser velocity, as they seek to provide ever more relevant content to their users be they employees of a corporate intranet or potential customers of an online store. The Coveo Relevance Maturity Model describes progressive levels of attainment for organizations seeking to provide the most relevant content and information to their customers, communities and staff. 5
Siloed Search Basic search captures users intent through them explicitly requesting what they are looking for, and saves time retrieving the requested information. At the simplest level, this is the bare minimum functionality everyone expects from a typical search box. This simplest form of search is confined to providing results that directly feature specific keyword terms in the search request, from a single source of data, or silo, and is of limited use. Level 0 - Federated Searches Federated search is an approach to information retrieval that enables simultaneous search of multiple disparate searchable sources. A user makes a single search request, typically in a search box, which is distributed to several search engines and databases included in the federation. The federated search then aggregates the collective results from the search engines and displays combined results to the user. The approach is highly resource-intensive, involving multiple search applications and business systems, all of which need maintenance and unique management skills. While federated search at least interrogates multiple data sources in parallel, it fails to rank combined result sets according to relevance. This frustrates users who are often overwhelmed with the increased volume of matching results. Like siloed search, it does little to provide information to the user with any kind of overall relevance weighting - hence Level 0. 6
Level 1 - Secured, Unified Ranking Like federated search, search results are consolidated from multiple sources. However, the results are unified and ranked, and fully compliant with access controls and security considerations. Search results are tuned and integrated. This eliminates having distinct relevance and ranking models from disparate search engines a key issue with federated search. Unified ranking applies a single rank profile across all of the results, ensuring the most relevant of those requested results surface to the top, irrespective of the data sources from which they originate. Level 2 - Content Navigation While unified ranking ensures the most relevant results are at or near the top of the list, users typically want to be able to filter and further dissect results. Configurable rich facets provide the very first step of personalization for users, so that they have natural classifications against which to narrow down the initial search results to navigate through content to more tailored, personalized results. Unified ranking combined with faceted navigation enables users to quickly drill into comprehensive results, to easily find the answers they are seeking. In effect, users are able to converse with the information. In this level, relevance is manually tuned according to the results achieved and revealed through analytics and the administrator s agenda. 7
Level 3 - Tunable Relevance Organizations at this level are providing information to users based upon systematic tuning of results through data models. These models are setup to take account of who each user is, what the task at hand is, and the context of the interaction or transaction; then rank the results accordingly with higher relevance tuned to each user s circumstances and situation. Relevance is tuned automatically, through applying query tuning models, rather than a complex, set of individual rules that quickly become unmanageable, and impossible to scale. Level 4 - Contextual Relevance Tunable relevance models are further refined by receiving additional contextual inputs at the point of the search query or related behavior. Example triggers could include variables that tune results based upon cues such as in-product requests, or task-specific requests. These in-context inputs trigger automatic queries or inject additional relevance clues to a search, to further tune the results. Organizations at this level are providing highly relevant information to users that is auto-tuned based on models and user-specific cues, to rank results uniquely to each individual, for a higher level of personalization. 8
Level 5 - Contextual Suggestions Further related information can be provided proactively to users based on an understanding of what they re trying to achieve. This information is not directly being sought, but is likely to help them. Examples include related content, products or services from entirely separate systems, enabling them to do more, learn more or buy more. Best-in-class organizations at this level are also able to recommend names of de facto expert individuals within their organization or community who are authoritative on the given topic. These contextually relevant suggestions are based upon automated analysis of topic authorship and the frequency that what these experts have written has proven helpful to others. In addition to authorship, conversations within email threads, communities and other areas may be included as clues to expertise, without revealing the originating email content. Level 6 - Self-Learning Predictive Recommendations At the highest level of relevance maturity, organizations are able to discern their users likely intent, by matching usage analytics and behavioral data, such as web click stream data, third-party purchase preference data, and making recommendations as to what content they are most likely to want or need. Machine learning algorithms auto-tune such recommendations and rankings, to maximize specific business outcomes such as shopping cart conversions, or support case deflections, enabling true one-to-one user engagement and upskilling. 9
Organizational Orientation to Relevance As organizations migrate through the Coveo Relevance Maturity Model, their orientation to information provisioning evolves. Those at the lowest levels of maturity (Level 0) are simply reactive, well below the waterline of what consumers expect in 2017 and beyond. For the purposes of the model, we will refer to these organizations as Relevance Laggards. Their priority should be to at least become more responsive to user needs by attaining Level 1. Organizations at Level 1 and 2 focus on at least being more responsive to information needs. With greater maturity at Levels 3 and 4, information is proactively offered up to customers or employees based upon the context of those individuals. The most mature organizations, those at Level 6, predict what information is most likely to be needed next - and these are the Leaders in the spectrum. 10
Reactive At Level 0 of the CRMM, content is provided to users purely on a request-driven basis, from one or multiple data sources. Relevance of results is limited, confined to basic keyword matching, and considered well below what modern customers and employees expect. Responsive At Levels 1 and 2, relevance is personal, and based on response to specific request made by each user. These levels range from providing unified ranked results from multiple data sources, to enabling filtering and narrowing of results using facets or categories. Proactive At Levels 3 and 4, relevance is contextual to each user s unique situation and circumstance. The relevance of results is automatically tuned in real time through techniques that weight results based upon implicit and explicit contextual factors. Predictive The upper levels of maturity, Levels 5 and 6 in the CRMM, are characterized by the most relevant information and content being anticipated, suggested and recommended. Relevance is driven by machine learning and artificial intelligence techniques that identify what content has been most useful for prior similar users or user journeys, and predicting what subsequent users will most likely need next. 11
Assessment: What s Measured, Gets Managed Every organization should assess their level of relevance maturity. By understanding the relevance of the information they are providing customers and employees, they are able to make policy, purchase and process decisions to improve. The CRMM provides a quick framework to assess your organization s level today, and what should be your next logical step to embrace. Coveo helps organizations progress through each stage of the model by becoming more relevant to internal and external audiences alike. Coveo provides professional guidance, implementation, and ongoing customer support, to ensure organizations see continuous improvement and increased relevance to everyone who engages with their brand. 12
About Coveo Coveo is the Relevance Company. We help organizations deliver the most relevant information, everywhere, every time. Recognized as the Most Visionary Leader in Enterprise Search and as a leader in Big Data Search and Knowledge Discovery, Coveo intelligently delivers knowledge and critical insights from across the entire IT ecosystem at scale, using intelligent search, analytics, and machine learning. From website intelligence and self-service applications to intelligent contact centers and intranets, Coveo partners with some of the world s largest enterprise technology ecosystems to provide a more intelligent, unified and engaging experience for customers and employees. Coveo currently has more than 1,500 activations in mid-to-large sized global organizations across multiple industries. For more information, please visit www.coveo.com and follow us on the Coveo blog, LinkedIn, Twitter, and YouTube. 13