What Next? From Search to an Insight Engine

Insight Engines are emergent leading you to the next best action.

Moving on from Enterprise Search to Insight Engines

Enterprise search is a term used to describe a mature technology sector that focuses on indexing content from many business knowledge repositories (File Shares, Salesforce, ServiceNow) and making it securely retrievable within an organization.  There have been many technology vendors ranging from large to niche, all offering to improve the visibility of business data (products, documents, knowledge) to help increase revenue, reduce cost and mitigate risk.

Historically these solutions are known for overly technical implementations focused on requirements that we often call “plumbing,” leaving users and their needs ‘out of scope.’  Tokenization, n-grams, relevancy algorithms, authentication (AuthN) and authorization (AuthZ) usually lead the discussion. 

Previously, Google made some progress in abstracting away the technical complexity with their now defuncted GSA product. However, all too often a ‘successful’ search project meant 1) wrestling (figuratively and technically) content out of SharePoint and (maybe) a File Share and then 2) getting that content to display on a search results page. 

I am the first to agree that the technical challenges of securely ingesting and serving content are critical parts of any search implementation, but, ignoring how users need to use this data is not the right way. 

A search solution cannot be a departmental solution, or solely focus on the technical aspects of content ingestion to measure success or failure.  The consideration of user experience needs to be front and center not ‘out of scope.’ Now more than ever you need to take the user, the context in which they are using the system and the expected outcomes into account when implementing a solution. 

Embrace the promise of Insight Engines

Insight engines as defined by Gartner, “augment search technology with artificial intelligence to deliver insights — in context and using various modalities — derived from the full range of enterprise content and data.”  Put another way these technologies make search an embedded experience that is aware of the context from which users are searching and has an idea of where you (should) want to go.  This integrated experience is why there is such excitement around the promise of technology.

Insight engines are a collection of component features rolled into platforms, rather than a single technology.  Some key subcomponent technologies to insight engines include Natural Language Processing (NLP), Machine Learning (ML), and Artificial Intelligence (AI). These are game-changing and impressive in their own right, but their impact is improved when you stay focused on how they can benefit the user.  One way we have found helpful to keep the focus on the user throughout the implementation is to keep in mind that the end user activity of searching is not what they are attempting to do. Search is a supporting activity that knowledge workers use to complete their tasks.  Keeping this in mind helps push search to be transactional and emergent.

Insight engines augment search technology with artificial intelligence to deliver insights — in context and using various modalities — derived from the full range of enterprise content and data.

Gartner

Successful search needs to consider external influences as well.  Your users are continually interacting with search in their personal lives.  Take a moment and consider the number of search interactions a user encounters when planning and taking a vacation.  These external interactions set expectations and engrain behavior in your users.  As a result of these interactions, delivering a list of results to your users is no longer a reasonable project implementation goal at least not if you are trying to provide real value and innovation.  Data need to be emergent.  Embracing this concept is foundational to delivering useful solutions.

Emergent Data = Recommendations

Data needs to be emergent.  Emergent data means using insight engines to deliver recommendations for the users next best action in the context of what they are already doing.  Elegant solutions do not require a user to go to another web browser and look for the data that they need to complete a task. The recommendations should present themselves in the normal flow of the task the user is working through. Recommending knowledge articles to a user creating a support ticket to help them self-service is the clearest example of this. 

Key Takeaways for moving to an Insight Engine

When planning your next implementation, you need to stay focused on the user and interactions that are valuable by:

  • Reviewing trends in consumer platforms (Google, Facebook, Fandango, GrubHub) to understand how users expect search to work
  • Understand your users’ journey or business workflow then map search interactions to these to allow the search to integrate into existing processes seamlessly

Insight engines are a game-changing technology that allow you to put the focus back on the user and deliver incredible experiences.  The promise of insight engines is impressive so much so that it can be challenging to get started. MC+A has insight engine implementation expertise across many verticals and use cases. We are always happy to discuss how we can help you get up and running with this technology.