Cognitive Computing - Part 3
Challenges and lessons in cognitive computing

Where should organizations start?

• Be clear on the strategic value of the investment to the enterprise and employees. Don’t be hypnotized by the features of the technology; focus on the area of biggest value and ROI. For knowledge management, initially that might be content curation or search.
• Define success in a compelling way relevant to the organization and the strategic intent. Paint a picture of how cognitive computing will unfold and of the kinds of value that management should expect to see and when. Set expectations for adoption curves accordingly.

How can organizations get ready for this new wave of technology?

• Do your best to head off concerns. Work with risk, privacy, security, legal and contracting, and seek their input from the beginning. Demonstrate to leaders that you have done your due diligence.
• Develop some compelling proof-of-concept examples by identifying needs and a pilot business group that is ripe for an application, has the resources to help fund the project and can produce measurable wins.
• Focus promotional efforts on newcomers (to the organization or to KM) since they have the least investment in maintaining any status quo. But at the same time, do not assume that new joiners or a certain demographic will buy into new technology just because it’s new technology. Unless the technology is invisible to employees (as in content curation), they must be sold on using it.
• Provide appropriate training and support so that users can be successful using the new technology. From a search perspective, even today’s best machine learning applications cannot curate as well as a human with experience in the knowledge domain in question. Set reasonable expectations and communicate them broadly so that users understand what is (and isn’t) feasible in terms of successful applications.

When and how should organizations start assessing the new tools and their impact?

Focus on a few key metrics that capture the value you expect to create. You don’t need to capture all the metrics; instead, decide which best communicate why and how cognitive computing will improve the business.

• Use surrogate measures (e.g., adoption and usage rates) to demonstrate progress early on.
• Evaluate satisfaction and capture whether users perceive value from the tools.

Predictions

Cognitive computing capabilities are maturing rapidly. The Advanced Working Group’s predictions for the next three years follow:

Search and content curation will be the first cognitive technology capabilities widely embraced by KM professionals. The journey toward more personalized search results is already well underway in the form of personalized portals and dashboards.

Costs will decline as cloud and app-based approaches become more widespread. That said, no one should oversell declining costs; applying cognitive tools to an organization’s proprietary data and knowledge will remain cost-prohibitive for many. Affordable programs or apps that do very specific things (e.g., scheduling meetings by scanning multiple calendars and considering the participants’ preferences) will precede more inexpensive, comprehensive implementations.

A digital sidekick is likely already in your pocket. With cognitive computing, it will become more savvy and capable. Employees will adopt intelligent personal assistants to augment their work just as they adopted mobile devices. As the data privacy and security concerns are addressed, IT departments will learn to integrate relevant behavioral data from those devices into the overall search, discovery and curation process.

Knowledge management has an opportunity to play a strategic role in cognitive technology adoption by partnering with IT, risk management and HR (among others). KM should collaborate with IT to implement the knowledge strategy and develop the technological roadmaps and infrastructure required. Risk management should be consulted to ensure that data security, privacy and legal issues are addressed, and HR must be part of the conversation to grapple with the workforce management implications of cognitive systems.

Although cognitive computing has tremendous potential to augment and enable knowledge work, its implementation should not distract KM programs from their core missions to retain critical knowledge and ensure its flow.

APQC’s advice is simple: Leverage the new capabilities where they align with an organization’s broader knowledge strategy and support them with the same compelling business case, training, change management, communications and measurement that all KM initiatives require. Do not become overly enamored with the technology for its own sake, and do not abandon tried-and-true KM approaches such as communities of practice in the mistaken belief that new computing capabilities will eliminate the need for traditional, person-to-person knowledge sharing.

Supercharging KM with cognitive computing

Employees face an exponentially increasing amount of data, and KM practitioners are in an excellent position to help workers navigate the mountain, filter the noise and leverage new opportunities for insights. Cognitive computing has the potential to provide structure to process that data in a manner that could supercharge KM activities. Cognitive computing will allow KM to deliver even more successfully on its promises to help employees leverage collective knowledge for their own professional growth and for the benefit of the organization and its customers.