Many years ago, I worked closely with Ford, one of the world’s largest corporations. It quickly became clear that, despite the vast amounts of data available within the company, it was almost impossible for any one person—whether in sales, marketing, product research, or senior leadership—to find answers from all of that information. And this wasn’t just a problem for a massive enterprise like Ford. Even at a company with just 50 employees, the same thing holds true.
And while this was true then, it remains true today.
When someone within a company needs information they assume must exist somewhere, though don’t know where, their first move is usually to ask anyone or everyone via Teams, Slack, Zoom, or Google Chat. This is amazingly inefficient as you never know if people will see the message, pay attention to the message and ultimately respond. And if they don’t, the individual can safely throw up their arms and at least say “I tried.”
The cost of this inefficiency is enormous. People waste hours every week looking for what should be easy to find. Sales teams are losing time they could be selling. Customer support teams could resolve issues faster and support more customers in a day. Marketers may commission new research when relevant data already exists. And subject matter experts (SMEs) are buried under a pile of questions they shouldn’t have to answer. Decision-makers at all levels could act more swiftly if they had quick access to the right data.
With all the technological advancements we’ve seen over the years, how can we still be in this position? Haven’t businesses invested massive sums in data storage and systems? Surely, by now, they’ve figured out search?
The problem is that data is messy—especially unstructured data like PowerPoints, PDFs, Word documents, audio files, video content, service tickets, and more. This data is hard to search and most systems just search keywords or filenames. And data doesn't just live in one place. It’s spread across platforms like SharePoint, Teams, Box, Google Drive, Salesforce, ServiceNow, Confluence, external vendors, and more. Each platform has its own interface and login, and only searches itself (and all too often, not very well).
And what about the magic that is Generative AI, surely this must solve for this? Isn’t this the job of an LLM? Unfortunately, no. The LLM’s are ultimately corporate victims to the challenges of enterprise search as well. How so? Generative AI, for all of its amazing ability is only as good as the data available to it. And if the right data isn’t readily available, it won’t know how to answer the prompt, it will pull from public data or worse, just make it up (i.e., hallucinate).
For example, have you experimented with Microsoft Copilot yet? Is it amazing?Sometimes, it is. When it has access to the data that allows it to answer the prompt it is pretty cool. Far too often, it is somewhere between meh to disappointing? It is stunning how many users are already giving up on Copilot and don’t think it works. Why is that? Because Copilot can’t succeed when enterprise search is broke, it needs access to the right data.
There is a solution, it ultimately falls under the label of Federated Search. Systems like Lucy have connections to 100+ enterprise technologies, they are multi-modal, work across data types, they are built to integrate in with enterprise security and they allow a search through a singular interface across all of this. Further, Lucy is an Answer Engine® and once she has searched, she provides an accurate answer with citation down to the page level or video / audio clip as to where she found it.
Here’s the thing: most businesses we talk to say, “You’ve never seen a company with data as poorly organized as ours.” This is a universal problem. If you tried to perfectly organize your data, it would take an enormous amount of time and resources—and by the time you were finished, the data would have piled up all over again. Plus, what works for one team may not work for others. One person's ideal taxonomy can easily turn into another person’s maze.
And that’s the beauty of a system like Lucy—you don’t need to perfectly organize your data for it to work. Lucy can handle your messy data, finding answers from wherever they’re stored, across all systems, without requiring you to spend time manually organizing it.
Knowledge management doesn’t have to be an unsolved problem. With technology like Lucy, businesses can finally put their data to work.
