Whenever we need to approach the internet, we use Google, Bing or DuckDuckGo. It is so natural that we don’t give it a second thought. We also don’t give a second thought to the fact that when we enter a support site, we dont think its search would yield any relevant results. Why is that?
Evolving the Insight Engine
As search evolves, the concept of a search engine rooted on the traditional ten blue links becomes less and less relevant. The days of the “I’m feeling lucky” button and links to click on are over. The problem with the technology internal search engines use, is that it is limited by its focus on query-based results and lack of semantic analysis. This calls for major changes in the way we think about search engines – or insight engines as they should be called – as they will continue to become more of our primary interface in web sites and intranet.
As we know, text is not the only thing that helps us find answers. In fact, in recent years the insight engine has been evolving far beyond simple text processing. We’re exploring video search and how that can benefit our lives as consumers and organizations.
The State of Insight Engines Today
Tens of millions of queries are asked to Google everyday and the search engine provides insights to users in the form of results. These results are designed to help users find information they are looking for. Despite this, as many as 20% of queries typed into search engines go unanswered. This poses a challenge to the design and implementation of insight engines. Can we build efficient, dynamic, user-friendly, and accurate insight engines that can provide more answers?
Insight Engines are changing the way we do research, uncover insights and make decisions. The new approach is more human centric and relies on creating a relationship with the user.
The Insight Engine is the next step in search and is built on the premise that we need to understand intent before providing answers. If you use a search engine, you’ll notice that you do not always see what you were searching for. This is because search engines still rely on people typing in keywords or using Boolean logic to find what they are looking for. There’s no context or meaning behind the words, which makes it difficult for a machine to provide relevant results. The Insight Engine can solve this by understanding the intent behind your question and providing results based on your goals and interests in mind.
Why the Current Approach to Insight Engines Won’t Cut It in the Future
Traditionally, insight engines have been built on keyword-based systems that try to understand consumers’ intent as they search for products across multiple channels like search engines, e-commerce sites, and social media. While this has worked well for the past 20 years, it will not continue to be effective in the future. With the emergence of conversational interfaces, the role of insight engines is shifting towards helping businesses create a superior user experience through a more natural interaction with consumers. Successful insight engines will need to evolve into a deeper understanding of what consumers want from the commerce experience and how to serve them within an increasingly competitive landscape.
A Better Approach to Delivering Answers is Needed
Conversational interfaces are rapidly changing the way we interact with devices, from voice assistants like Siri and Alexa to chatbots on messaging apps like Facebook Messenger and WeChat. These interfaces allow us to express our intentions and needs in conversational language rather than search keywords. Rather than instructing Siri to “search for an iPad Pro store” or asking Alexa if Amazon sells shoes, we can simply ask “Where can I buy an iPad?” or inquire “Do you sell shoes?” This shift away from keyword-based search may seem minor at first, but it has major implications for commerce businesses of all kinds.
A better approach to delivering direct-to-answer solutions is needed. The power of cognitive computing allows us to move beyond just generating keyword-based results expanded a bit with synonyms, toward a new frontier of training models on vast amounts of data available on the Internet, in order to deliver actionable insights.
To get there, we need more powerful computing capabilities and new algorithms that will allow us to do things like learn from knowledge already in our data repositories, reason over unstructured content in order to make accurate inferences, understand sentiment (i.e., whether content is positive or negative), and extract entities from a variety of content sources (e.g., video, audio, text).
We Are Evolving Insight Engines to Create the Future of Search
As we look to the future, it is imperative that there is a focus on a new type of user interface for search. The experience of searching should evolve from a paradigm of searching for keywords to a paradigm of question answering.
Korra delivers direct answers faster with a new approach to search with intelligence and analytics.