Many thanks to Balajii Chandarasikharan for suggesting this topic.
When working with partners or with a tiered support model of any kind, it is useful to monitor the volume of escalations, or, more properly, handoffs from tier 1 to tier 2 (or from partners to the vendor. After all, handoffs suggest that the first-tier support engineers do not have the technical knowledge required to resolve the issue, right?
Not really. What matters is not so much the raw volume of handoffs but rather the volume (or ratio) of unwarranted handoffs. Let us review:
- A valid handoff is one that cannot be avoided, for instance a bug, an enhancement request, or a tough question that was not documented prior to the handoff.
- An invalid handoff is one that should not have been needed: a question for which an answer was documented in the knowledge base, or some other repository.
The individual who handles the question is usually able to determine whether the handoff was valid — although to be sure there are gray areas: perhaps something was documented in the knowledge base, but in such an obscure manner that the handoff really could not be avoided. Still, if a new bug was created as a result of the handoff, or a new knowledge base document, the handoff was clearly warranted.
- Measure the volume and ratio of handoffs, but don’t set targets on them, especially at the individual level. If someone hands off 20% of their cases, it could be absolutely alright if virtually all handoffs are related to bugs. Don’t penalize the poor soul!
- Seriously consider measuring invalid handoffs. We are all human: a small number of invalid handoffs is fine, but significant amounts, especially sustained, are cause for alerts (and possibly penalties for partners).