Search doesn’t matter. Answers do.
Search doesn’t matter. Answers do.
No matter how many levels of complexity a search engine contains, drill-down abilities don’t define its worth. In the end, the best search result is the one that gives you the answer you need. Insights organizations usually store all their research on intranet portals and share drives. At worst, users need to know the exact name or ID of the project they need to find it; at best, they’re organized by taxonomies with an in-document search.
But this makes answering business questions difficult. For example, finding the answer to a question like, “What do we know about organic ingredients?” could take hours of effort, opening and reading through hundreds of PowerPoints and PDF files. (Of course, Market Logic’s insights portals made this much easier, by manually extracting findings to deliver a one-click search result.)
The changing AI landscape
Today, AI is changing the landscape by understanding the intent behind your questions, and reading through all content to find the most relevant answers for you. Think of a portal that recognizes what you need as soon as you ask for it. One that delivers not just the best report on the topic, but summarized content from inside the document that cuts straight to the chase by directly answering your question.
At the next generation market insights platform launch, Mario Lenz, Senior Vice President, Product Management, explained the thinking behind the new search engine, pointing out that “Search doesn’t matter. Answers do.”
The new search experience
Inside the new market insights platform, Market Logic has completely rebuilt the entire search experience with state-of-the-art AI to ensure the machine understands the intent of each user. The platform connects all data, structured and unstructured, in a client-specific knowledge graph, which continuously learns from new information and business actions.
The AI powering the search uses different strategies depending on questions. If a user asks a question like “what scents are trending today,” then results will focus heavily on social data and syndicated research; if the user needs to know their brand’s market share in a specific region, top answers will come from their POS data.
A truly semantic search
The engine executes a truly semantic search, as it recognizes relationships between words and phrases, not just acronyms and synonyms. Relevant content is proposed as search terms are entered, and Mario explained that “the engine delivers the best direct answer to your question from the most reliable source.”
When a search term or question is entered, the engine collects related files and summarizes information from all of them, without any repetition. These automatically extracted findings are pulled from project documents and presented to users “with no manual work required, whatsoever.”
When marketers go to a portal to find information, their end goal is an actionable insight. A system that can give direct answers has obvious benefits over one that only returns document results. When your platform comprehends the semantics of a question and uses all your research and data to respond, time savings are exponential.
The next gen Market Insights Platform
At the launch, Mario shared a personal example to illustrate this: a short while ago, he’d been telling his son about the new platform and his test work with an illustrative set of research data on shampoo. His son, an avid tennis player, asked if the platform could recommend a new shampoo for use after training. So, Mario logged in and asked, “What haircare scents do millennial men prefer after sport?” The system immediately returned the best answer – curiously enough: “Millennial men prefer floral and fruity scents.”
In short, the platform proved that it understood Mario’s intent in asking a highly specific question: he wanted to know of particular haircare scents young men are looking for, not just reports and findings on millennial men, shampoo and fragrance. The knowledge graph within the platform had instantly looked at all connected data to pull a direct answer from the best sources it contained. To support the assertion, it had also returned all relevant projects, pulling out key findings to create an automatic summary in the sidebar.
The combination of all available data, together with a platform that understands user needs, means we’re moving from an era of asking questions to an era of instant, actionable answers. The new search experience gives users what they need when they need it.