This blog is a follow-up of a previous blog where we explained the challenges of most support organizations preparing for a shift-left initiative. As explained in that blog, the traditional tiered support model is no longer effective after a successful shift left and employees skills are mostly under-utilized, causing them to become disengaged with the company they work for.
It should be clear that there is no easy fix for these problems. After all, they originate from more than 30 years of accepted practice and structure of support. Changing this require a radical change of perspective. So let’s ditch all the legacy and try to reinvent a support organization. What would that look like?
First, we would view the entire support organization as a single team with a common goal: solving issues as fast as possible to restore the productivity of the customer. Individual team members have very diverse skills, talents and interests. Of course we want to make the best possible use of this diversity. We would also recognize that the required skills to solve a certain issue strongly depend on the characteristics of the issue itself. So ideally, each issue is handled by the people with the all the skills required to solve it. It could be a single person, it could be a large group and all variations inbetween.
Handovers of issues disrupt the solving process, inhibit learning and knowledge transfer and simply takes longer. So the person to whom the issue is assigned should be the person that is most likely to solve it. He or she will own the issue from registration to solution and is responsible for its progress. They can invite others to assist but will never lose sight on the resolution When the issue is resolved the owner has end to end knowledge of the issue, the solution and all the steps in-between. And probably learned something in the process.
Because people enjoy handling issues that match their interests, skills and talents the most they should be able to choose the issues to work on and volunteer to help others. They pick the issues for which their assistance would be helpful from a pool of open issues. This empowers them to work on their strengths and allows them to contribute more than just knowledge like testing, researching, communication, out-of-the-box thinking or coordination. This making the work more enjoyable and rewarding.
This dynamic collaboration model requires knowledge workers to interact with others and identify the required skills and knowledge for resolving the issue at hand. This requires a broader understanding of issues, the business of the customer and your colleagues. It also requires the ability to collaborate across boundaries like teams, knowledge, language, culture and business. This means that knowledge workers need to evolve into a specialist and a generalist at the same time, a so-called T-shaped individual.
This completely different way of organizing support is called Intelligent Swarming and is developed by the Consortium for Service Innovation and its members. This page contains links to various case studies that show how successful this methodology can be. However, experience so far also shows that this is not for everyone. A successful adoption of Swarming requires a substantial organizational change. A culture of control and (micro) management needs to evolve into leadership, enablement and trust. Sceptics will question the principle of opt-in, where knowledge workers volunteer to help rather then getting issues assigned. Although the methodology provides a mechanism for detecting issues that are not picked op by anyone, experience learned that this is rarely necessary. Teams are remarkable capable of handling these responsibilities without any outside interference.
The objective of assigning issues to the people that are most likely to solve them of is often misinterpreted as ‘just assign everything to your most experienced and knowledgable people because they will likely be able to solve it’. These people will then be swamped with requests while the rest of the knowledge workers are looking for something to do. The algorithm needs to be a bit more intelligent of course and should look for someone that is just qualified enough to solve the issue. It can also take labour costs into account: if two people with similar qualifications are available choose the individual that costs less. Call it a ‘minimal effective dose’ principle. Skills and the ability to resolve complex issues will also develop much faster in a Swarming environment.
A bigger challenge is actually building the algorithm itself. You need to build and maintain people profiles that take various aspects like knowledge, experience, skills, certifications, interests and labour costs into account. Another part of the profile consists of a ‘digital reputation’: the requests you solved (or helped to solve), the tasks you worked on, the knowledge articles you created or improved and customer feedback your articles and resolutions. This information is extremely helpful in determining when you would be the right person to ask for help.
Maintaining people profiles manually is possible and is often the only viable option when starting with Intelligent Swarming. But it is clear that there is a big opportunity for machine leaning and artificial intelligence here. Organizations that successfully use swarming often did substantial investments in automating the creation and maintenance of people profiles and reputation. This type of functionality is probably the biggest opportunity for vendors of service management solutions to stand out in a very crowded market.
“Time to move Artifical Intelligence out of the buzzword stage”
It would be great if these vendors recognized the power and necessity of new ways to organize support like Intelligent Swarming. It would also provide a clear direction for the development and application of machine learning and AI in these tools and move beyond the buzzword stage. You don’t have to wait for it, however. You could start today and be prepared for your future shift-left greatness.
