Video Script (Video Length: 5:41)
This is the sixth and final video in a series that, in its entirety, provides a brief primer on Collaborative Intelligence, its applications, and its implications for business leaders.
In the previous video, I presented five principles that, if adhered to, will improve the results you see from your CI efforts.
This video will provide a high-level process that can guide you as you introduce CI to your organization. It adheres to the five principles.
ID Candidate Workflows
The first thing you must do is identify prospective workflows that could benefit from the application of collaborative intelligence.
The most promising candidates will exhibit the following four characteristics:
They are recurring. Working on seldom-used processes may reduce risk, but such an approach also reduces organizational impact and limits direct experience of the redesigned processes’ benefits.
Remember: initial CI projects should demonstrate valuable benefits and build organizational support for additional efforts. Improving seldom-used processes rarely accomplishes both goals.
Designing workflows that are new, or tackling previously unsolvable problems, should wait until your organization has some experience with CI. Such projects are typically larger and more challenging than those that redesign some or all of an existing process. Such projects should be tackled after you’ve built some CI muscle.
The next two characteristics are related. The candidate processes should be time-consuming, resulting in reduced throughput and inefficient utilization of expensive or limited resources.
As I mentioned in the first video, collaborative intelligence aims to redesign work in ways that assign appropriate tasks to humans and intelligent machines. Much current work requires that senior managers, subject matter experts, and highly skilled staff complete low-value and repetitive tasks and participate in many low-value meetings.
Such tasks are better assigned to intelligent machines, and as many meetings as possible should be rendered unnecessary in CI-enabled workflows. The improved work design increases throughput and enables highly-paid and highly-skilled employees to do more high-value work.
Candidate workflows should contain several perfunctory tasks that are completed currently by many different people across the organization. In many cases, the resulting work products are highly variable because their quality depends on the skills and work habits of the specific individuals completing the tasks.
The improvement in natural language processing allows intelligent machines to complete such work more quickly, consistently, and without fail, freeing valuable staff to focus on higher-value activities.
Once several promising candidates have been identified, select two or three for redesign in parallel. Multiple redesigns expose the organization to a more diverse set of challenges, opportunities, and technologies. Parallel efforts also provide opportunities for collaboration across teams, and more rapid learning about how CI can help your organization.
ID / Baseline Performance Metrics
Once the candidates have been selected, you must baseline current performance, against which you will compare future performance. Like any technology, CI should be adopted only when it provides valuable business benefits. Performance improvements cannot be measured if current performance is unknown or only discussed anecdotally.
Three categories of metrics should be used to assess current and future performance. Two are well-known—quantitative and qualitative—but a third is equally important: experiential, which includes metrics such as changes in quality of work life and changes in perception about CI.
Since one goal of CI is to create work processes that better utilize human capabilities, capturing this experiential data is very important. It helps leaders:
build support for additional CI efforts,
identify issues that need to be addressed prior to initiating this additional work, and
set realistic expectations about CI’s capabilities and its organizational impact.
Iteratively Design / Refine CI-Enabled Process, Execute / Measure New Process
Once metrics have been defined and baseline measures recorded, the redesign work can begin in earnest.
CI initiatives should be designed and executed as a series of business experiments that enable your organization to learn quickly what works and what doesn’t, adapting along the way until a CI-enabled process is outperforming the baseline process. If such outperformance is not attainable, it is important to understand why, so that these barriers can be shared with other teams.
It is also helpful to record executives’ and team members’ questions and concerns prior to beginning each effort, so that, once concluded, any answers or insights can be captured and shared across the organization.
Share Results and Lessons Learned
Once each initiative is completed, the experiences and lessons learned should be shared, along with comparative performance data. This is meant to be a pragmatic exercise that offers a clear-headed review, not a propaganda session designed to oversell the value of CI.
The goal is to help the organization identify the most promising areas of application right now and those that may be a bit too ambitious given the current capabilities of CI technologies and the organization’s experience and expertise in applying CI.
Following the guidelines from the previous video and the high-level process described in this video should enable your organization to get off on the right foot with collaborative intelligence.
Thank you for watching.
Share this post