What is the difference between deflection and demand elimination?
Deflection moves a contact to a cheaper channel — self-service, chatbot, knowledge base. Elimination removes the reason the contact happened at all. Deflecting a contact caused by a wrong bill still means the bill was wrong, the customer was still inconvenienced, and you are still paying — just less per contact.
| Deflection | Elimination | |
|---|---|---|
| What changes | The channel the contact arrives on | Whether the contact happens |
| Who acts | Contact centre and digital teams | Whoever owns the upstream defect |
| Effect on cost | Lower cost per contact | Cost removed entirely |
| Effect on the customer | Still inconvenienced, now self-serving | Never inconvenienced |
| Ceiling | Limited by what self-service can handle | Limited by what can be fixed upstream |
| Failure mode | Making it harder to reach you and calling it success | Chasing causes that can’t actually be fixed |
Deflection isn’t wrong
For genuine value demand, self-service is often what the customer wants — no queue, no hold music, done at midnight. Deflection is a reasonable strategy for demand that should exist.
The problem is applying it to failure demand. A customer whose parcel is missing being routed into a chatbot has been failed twice. And because the deflected contact disappears from the voice statistics, the underlying defect gets quieter without getting smaller — which makes it less likely to ever be fixed.
The order that works
Classify first. Eliminate what shouldn’t exist. Deflect what should exist but doesn’t need a person. Route the rest well. Doing this in the wrong order means automating your defects instead of removing them.