In short
Usually because traffic that was never legitimate has stopped, or because senders have temporarily rerouted while they assess the new controls. A drop in volume alongside stable or rising revenue is a successful deployment. A drop in both volume and revenue, with subscriber complaints, indicates over-blocking that needs immediate review.
The expected drop
Some decline is the deployment working. Artificially inflated traffic stops because the fabricated destinations are now identified. Spam and smishing campaigns move to networks with weaker controls. Bypass operators pause while they test which techniques still get through.
This category of decline should coincide with stable or improving revenue, because the volume that disappeared was never being billed in the first place.
The temporary drop
A second category is displacement rather than elimination. Aggregators that were terminating unofficially reroute to another path while they decide whether to sign an agreement. Traffic dips, then returns — either through the front door under a commercial agreement, or through a new bypass route that needs detecting.
Watching where displaced volume reappears is one of the more useful activities in the weeks after go-live, and a good argument for keeping route testing running continuously rather than as a one-off audit.
The drop that signals a problem
Over-blocking looks different, and the distinguishing indicators are unambiguous:
- Revenue falls alongside volume, rather than holding steady
- Complaints arrive from named enterprises, particularly banks, about failed OTP delivery
- Blocked-traffic reports show high volumes from registered senders rather than unknown ones
- Delivery rates fall for accounts with existing commercial agreements
Why the baseline is what saves you
None of these judgements are possible without a pre-deployment baseline. An operator that enabled enforcement immediately has no way to say whether a 20% volume decline is recovered bypass or blocked banking traffic, and will usually end up relaxing controls indiscriminately to stop the complaints.
Monitoring mode exists precisely so that this conversation can be had with data. It is also why enforcement is introduced progressively, by category and by route, rather than everywhere at once.