Each one requires the user to use that system, nearly exclusively, for knowledge—creation, sharing, reusing. But knowledge is inherent in everything knowledge workers do, and the last thing they want or need is yet another system to learn, go to, create stuff in, find stuff in, and leave, so they can get back to work. Experience (and research) proves that most just won’t use it.
A number of the KM systems reviewed include microblogs/collaboration apps. And yet, everyone tugs at the knowledge worker to use their system of choice as the corporate standard. The marketing and sales teams, and maybe customer service, want everyone to use Salesforce Chatter. IT wants everyone to use Yammer. But everyone really wants to use Twitter and LinkedIn, and maybe Google Docs. Now here comes KM with yet another system. What happens is that different people will use different systems.
A colleague shared this stat with me, from the Worldwide Intranet Challenge: Approximately 90% of staff do not regularly contribute to the intranet, [and] more than 50% never do. That’s because people work in the systems that they work in, they don’t go somewhere else, like the Intranet, to “create knowledge.”
Knowledge creation is simply a by-product of work (which is done by engaging with other knowledge, information, and people). We need to capture it in the flow of work, in the systems people choose to use, and make it broadly available where they work. There is certainly a place for curated knowledge in a system of record, but it is part of the overall knowledge and information ecosystem, not the exclusive holder of it.
Knowledge sharing must also become part of our work-flow, when it’s relevant to us and our colleagues, and via the systems we choose. It should simply appear related to and based on what we are doing. A law-journal article published this week about the future of knowledge management termed this capability Artificial Intelligence (AI). I’m not sure it’s reached that level yet; it’s called Predictive Knowledge or Predictive Intelligence, depending on the group using it. I don’t blame the person who was quoted, because the experience does seem like magic. In reality what happens is that the user and his or her context become the query, so no search is conducted overtly, it’s all done in the background, invisible to the user. The knowledge worker simply sees relevant content and experts recommended by the system, based on the work she or he is doing.
It sounds simple, but there are several components necessary to enable this type of Predictive Knowledge. Key to user adoption, and setting the stage for program success, is an interface layer that may be configured to reflect how people work, and appear within the systems they work in most. Rather than making people go to a knowledge system to do “knowledge work,” the knowledge comes to them, in the places they prefer, in the flow of their work.
Other requirements include broad connectivity, a virtual information integration (unified index), analytics, and a relevance engine that is open to administrators to tune, and takes in clues about people’s interactions with data to make it ever more relevant.
Sounds like magic, but today it’s reality.
Where do you see the biggest impact of Predictive Knowledge?