One of the criteria for entering Phase 3 is that enough content has been collected in the knowledge base that customers have a better than 50% probability of finding something helpful through self-service. The indicator for this is that the internal article reuse rate is equal to or greater than the create rate. In Phase 2 we used article reuse as the indicator for which articles should be published to self-service. As we move through Phase 3, we want to move away from using reuse counts as the deciding factor on what gets published for self-service, to relying on the judgment of the KCS Publishers. The goal? 90% of what is known internally, and usable by the customer, will be published to self-service at the time we know it (at or before case closure). It is a bold goal, and more information can be found in this case study.
We want to get as much of what we know to the self-service model as quickly as possible. This is because in mature KCS environments with effective self-service models, we see the reuse patterns of articles in self-service is very different from the internal reuse patterns. Most support organizations only see a very small percent (typically less than 3%) of the total customer demand for support (for more on this topic see the “Demand Based View of Support” presentation at www.thekcsacademy.net/kcs). This means customers don’t open an incident for the majority of the exceptions or issues they encounter. Data from Consortium for Service Innovation members shows that customers will use a good self-service mechanism ten times more often than they will use the assisted support model (support center). Internal reuse is a reasonable short-term indicator of the value of an article, but in the long term, we want to enable customers to access most of what we know because they will often reuse articles, or solve issues, through self-service for which they would never have opened an incident.
While many issues resolved by customers using self-service will be those for which they would not have opened an incident, some of their self-service success will be on issues for which they would have opened an incident. This means the nature of the work coming into the assisted model (support center) will shift over time. A higher percentage of incoming incidents will be new, as many of the known issues are now solved through self-service. Knowledge workers will be spending more of their time solving new issues, and this results in a dynamic that may not have been expected.
As self-service becomes more effective, many of the support measures will move, in what traditional terms would be “the wrong direction." Time to resolution and cost per incident will increase due to the fact that we have removed many of the known or easy issues from the work mix, and it takes longer to solve new issues. This is reflected in an increase in the new vs known ratio, and this ratio is an excellent indicator of the effectiveness of your self-service model.
Five Key Enablers for Self-Service Success
More detail on the enablers for self-service success can be found in the KCS v6 Practices Guide, but as a reminder:
- Findability - The content must be in the context of those who are using self-service (KCS addresses this)
- Completeness - Most of what we have learned internally must be available to self-service quickly; 90/0 (KCS addresses this)
- Access - Making it easy to access self-service for the users (integration of self-service into the application interface)
- Navigation - Support browsing and searching, and offer easy transition from self-service to assisted (no dead ends)
- Marketing - A plan to encourage use of self-service