Messenger engagement dropped 20% in two weeks among advertisers. Diagnose the issue

Problem Statement Description

Product context: Meta is a social technology company; its products include Facebook, Instagram, WhatsApp, Messenger, Threads, Quest, creator tools, and ads.

Meta has observed a 20% drop in Messenger engagement over the last two weeks among advertisers. In this context, “advertisers” may include businesses, agencies, and creators using Messenger to communicate with customers, manage leads, respond to ad-driven conversations, or support commerce-related interactions. Your task is to diagnose what may have caused the decline and determine how the team should investigate it.

This is a root-cause analysis problem, not a product redesign question. You should frame the anomaly clearly, define what “engagement” could mean, identify whether the drop is real or instrumentation-driven, and break down the issue across relevant advertiser cohorts, surfaces, geographies, devices, campaign types, and Messenger workflows.

The investigation should account for Meta’s broader ecosystem, including ads delivery, click-to-message campaigns, business inbox tooling, notifications, account integrity systems, customer response flows, and possible external market or competitor effects. You should distinguish between product, technical, measurement, policy, and demand-side explanations before deciding what evidence would confirm or reject each hypothesis.

The experience should consider:

- How to define the affected engagement metric, including numerator, denominator, time window, and advertiser eligibility

- Whether the decline is isolated to Messenger or connected to ads products, business messaging, notifications, or account access

- Segmentation by advertiser size, vertical, region, device, platform version, campaign objective, and new versus existing advertisers

- Instrumentation checks for logging changes, event schema updates, data pipeline delays, bot filtering, or attribution changes

- Product and technical hypotheses such as message delivery issues, notification failures, inbox UI changes, latency, crashes, or API disruptions

- Business and policy hypotheses such as ad auction changes, moderation enforcement, spam controls, privacy changes, or seasonality

- Evidence needed to prioritize the most likely causes and separate correlation from causation

- Immediate mitigation, stakeholder communication, monitoring, and prevention steps once the cause is identified

Your goal is to demonstrate a structured diagnostic approach that narrows a broad engagement decline into testable hypotheses, validates the reliability of the data, identifies the impacted advertiser experience, and recommends an evidence-driven path toward mitigation and prevention without jumping prematurely to a single explanation.