Over the previous few months, my colleague Matt Mullen and I’ve been taking briefings, researching educational and business sources, and brainstorming a brand new space we name “work intelligence.” It’ s not one thing we’ve invented: It’ s a motion we’ve watched emerge through the previous couple of years. In easy phrases, we’re speaking about utilizing AI to converge information sources and insights from staff, clients, operations, and processes right into a unified complete, the objective being to drive higher enterprise determination making. In case you are as geeky as we’re, it’s fascinating stuff. Not solely is there an rising market, it is going to probably be a giant one. However underlying all our analysis and discussions has been the uncomfortable actuality that, though work intelligence makes excellent sense—the expertise is accessible at present to make it work—it doesn’t basically work. Or, to be extra particular, it really works, however not in addition to its champions would have you ever imagine.
Shortcomings in KM problem-solving largely because of a mismatch between what the expertise guarantees and what it may possibly truly ship are recurring themes on this column, however 20 years in the past, within the dot-com period, the issue wasn’t a scarcity of dedication or ambition; the expertise out there then simply wasn’t excellent. Quick-forward to 2023, and the expertise out there is mind-blowingly good, dependable, and, generally, inexpensive and accessible. The issue at present isn’t the shortcomings of the expertise. As an alternative, it’s the lack of a constant method to utilizing it, together with poor high quality information, and the assumption that extra information and processing energy will resolve any kinks within the system.
Let’s begin with a constant method. In the event you haven’t set clear enterprise targets and outlined a short-, mid-, and long-term path to reaching these targets, the chances are that you’ll fail. Then there may be the difficulty of information. Know-how, notably AI, wants clear and correct information, however seldom will get fed that diploma of information high quality. And this is the crucial level: It by no means will. Belief me, initially of each venture, the necessity to clear and preserve information is harassed, however it’s a message companies do not wish to hear. Give it some thought this fashion: Information from an IoT gadget might be fairly darn correct. It can learn the temperature, location, humidity, and if a tool is switched on or off. It gained’t be excellent, however it’s most likely ok. However examine IoT information with the hundreds of thousands of paperwork saved in SharePoint or an HR system. The standard, accuracy, and consistency of the information saved there’ll likley to be comparatively woeful.
Some instruments (sometimes AI-based) can enhance information high quality, however they’ll by no means be 100% correct and all the time lack human context. However that doesn’ t imply you shouldn’t use AI or a new-fangled work intelligence system sooner or later. Removed from it. Technological advances are vital and may convey large advantages, however solely so long as you perceive that they’ll advise, increase, and assist, however not substitute, you. Even probably the most superior AI techniques, fed with the highest-quality information, are surprisingly restricted in what they’ll do. They do some issues exceptionally nicely, higher than any human. However in the case of operating your enterprise, making crucial choices, planning, and reworking, they need to all the time be in a supporting, not the main, position.