Back to blog SMS Operations

Escalating A2P SMS Incidents: Levels, Owners, and Evidence

A practical guide to handing off A2P SMS delivery incidents with context, evidence, and clear limits—without mistaking a DLR for confirmed receipt.

Operations teams review evidence and coordinate the escalation of an A2P SMS incident

Why A2P incident handoffs get stuck

An incident may pass through the team integrating the application, a connectivity provider, an aggregator and, where applicable, a mobile network operator. The next team cannot investigate effectively if it receives only a description such as “SMS messages aren’t arriving.” It needs to know what traffic is affected, during which time window, and what checks have been performed.

This article covers handoffs between teams and organizations, not how to detect an anomaly or reconcile statuses on a platform. The goal is to turn an initial observation into an investigable question while keeping facts, statements received from others, and hypotheses separate.

  • A message status alone does not describe an incident’s scope or prove its cause.
  • Each organization may have different visibility into events on its part of the route.
  • Shared information should be sufficient for the investigation and limited to what is necessary.
Why A2P incident handoffs get stuck

Define levels by impact, scope, and urgency

The available evidence does not establish a universal severity scale for A2P SMS incidents. Each organization should agree on its own levels and document what they mean. Consider service impact, observed scope, and operational urgency together rather than assigning priority based only on a delivery status.

Before escalating a case, describe who or what is affected based on the information available—for example, a specific flow or sender, a destination, or a defined set of samples. Do not generalize to an entire country, operator, or route based on isolated observations.

  • Impact: which legitimate process is affected and what consequence has been confirmed.
  • Scope: which traffic, destinations, and time period the evidence covers; also state what is not yet known.
  • Urgency: what decision or response the next team needs, and by when, under the applicable operational agreement.
  • Priority: record the criteria used and review the classification if the scope changes or new evidence emerges.
Define levels by impact, scope, and urgency

Separate facts, statements, and hypotheses

A clear handoff distinguishes what was directly observed from what another party reported and what the team suspects. This prevents a provisional explanation from becoming an attributed cause through repetition, without confirmation.

Describe each claim with its source and scope. “The system recorded an event with this status at this time” is different from “the provider said the message was accepted”; both are different from “the cause may be with the operator.”

  • Observed fact: data recorded by an identifiable tool or check.
  • Statement received: information attributed to the team or system that provided it.
  • Hypothesis: an explanation that has not been validated; include what evidence could confirm or rule it out.
  • Unknown: an aspect for which there is not yet enough evidence.

Prepare a minimum handoff package

This guide does not establish a mandatory set of fields applicable to all providers or operators. As a coordination practice, agree on a common format that helps locate samples and understand the time window under investigation. Use references the recipient can correlate and clarify which system they belong to; an internal identifier may not be visible outside your organization.

Share minimized samples. Avoid including the full message body or personal data unless they are necessary for the investigation. Apply the privacy, security, and retention rules in force at the organizations involved.

  • Summary of the symptom, known impact, and specific question for the recipient.
  • Time window under investigation, including date, time, and time zone.
  • Destination in a normalized format agreed by the parties; E.164 may be used as a reference for representing international numbers, where applicable.
  • Sender or originating identifier used, and relevant context about the legitimate traffic.
  • References available in each system and how they relate to one another; do not assume they are shared across organizations.
  • Observed events, their timestamps, and the system that recorded them; attach a representative sample and explain how it was selected.
  • Checks already performed, their results, and visibility limitations.
  • Agreed priority level, case owner, and next update point.

Assign an owner and one question for each team

The case owner coordinates the work: maintains the current summary, records decisions, requests missing data, and communicates updates. This does not mean they know every technical segment or can confirm events outside their area of responsibility.

Phrase a question the receiving team can answer based on its visibility. Avoid broad requests such as “investigate delivery.” A well-defined question helps the recipient confirm what they can check, which references they need, and what they cannot verify.

  • For the internal team: What events did our systems record for these references and this time window?
  • For the provider: Which events and responses can you confirm for the identified samples, and what additional references do you need?
  • For the operator, if escalating through the agreed channel: Can you review the case using the available data and clarify which part of the route is within your visibility?
  • Record who is expected to respond, what information is expected, and when the case will be reviewed, under the applicable agreements.

Escalate in stages and respect visibility limits

Start by validating data under your organization’s control and completing the handoff package. Then direct the inquiry to the provider with whom the appropriate operational channel exists. An operator’s involvement depends on the applicable agreements, processes, and available visibility; do not assume a sender can escalate directly or that every party sees the entire route.

At each stage, preserve the original response and note what is confirmed, what remains pending, and what the next team needs. If references or context are missing, complete the request before forwarding it as though it were a conclusion.

  • Internal stage: validate identifiers, time window, destination, sender, and available records.
  • Provider stage: share the minimized sample and request a narrowly scoped check.
  • Operator stage: use the channel provided for the commercial or operational relationship, where applicable.
  • At every stage: document visibility limits and avoid attributing a cause to a segment without evidence.

Treat DLRs as limited evidence

The information available for this guide does not define the meaning of any particular DLR or establish that a given status confirms receipt on the handset. Therefore, preserve the status and its source as an observation; do not automatically turn it into proof of delivery to the user, reading, or a cause of failure.

If events are delayed, contradictory, or incomplete, record their timestamps and source systems, note the discrepancy, and ask the team able to consult that segment for clarification. Do not resolve a contradiction by choosing one status without explaining the criteria or uncertainty.

  • Distinguish a received event from any independent check of receipt on the handset.
  • Note when the DLR was observed and in which system; do not confuse that time with the send time.
  • If two records differ, preserve both and ask for clarification about their scope.
  • Do not infer a technical cause from a DLR without additional evidence.

Update, close, or reopen with traceability

Define in advance how updates are recorded under internal procedures and agreements between organizations. Each significant change should state what new evidence emerged, what interpretation changed, and who made the decision.

The available evidence does not establish universal closure or reopening criteria. Set them in the operational process—for example, what response is sufficient to close the investigation, which uncertainties remain documented, and what new evidence would justify reopening it. Closing a case should not erase any remaining limitations.

  • Maintain an up-to-date summary, links or references to evidence, and a timeline of decisions.
  • When closing, record the confirmed outcome, any hypotheses ruled out or still pending, and the limits of the conclusion.
  • When reopening, cite the new evidence and explain why it changes the analysis.
  • Retain data only for the period permitted by applicable policies and obligations.
FAQ

Frequently asked questions

Does a DLR confirm that the user received the SMS?

Do not assume so without knowing what the status means and what the system that generated it can see. In the handoff, record the DLR as an observed event and distinguish it from independent confirmation of receipt on the handset.

What information should I include when escalating an incident?

Include a description of the symptom, known impact and scope, the time window with time zone, destination, sender, correlatable references, events and their sources, minimized samples, checks performed, and a specific question for the receiving team. Agree on the format with the parties involved.

When should the operator be involved?

When the investigation requires reviewing a segment that may be within the operator’s visibility and there is an appropriate operational channel for doing so. Whether escalation is possible and what data is required depend on the applicable agreements and processes.

How should I handle contradictory responses from different teams?

Preserve both responses with their sources, timestamps, and scope. Ask each party to clarify which references or events it was able to check, and leave the cause unconfirmed until the evidence resolves the discrepancy.

Is there a universal severity scale or set of closure criteria?

The available evidence does not support a universal scale or a single set of criteria. Define levels, owners, and closure and reopening conditions in the operational procedures of the organizations involved.

Sources consulted

  1. 3GPP specifications by series3GPP
  2. ITU-T Recommendation E.164International Telecommunication Union
  3. GSMA networks resourcesGSMA
  4. A2P SMS: qué es y cómo se diferencia de P2P1NCE
  5. El mercado de mensajería A2P: problemáticas, retos y oportunidadesTelefónica / Analysys Mason