Retention for advertisers declined in YouTube. What is your analysis plan

Problem Statement Description

Product context: Google is a consumer technology, ads, AI, and cloud company; its products include Search, YouTube, Android, Maps, Gmail, Chrome, Google Play, Workspace, and Google Cloud. YouTube is Google's video platform; its products include long-form video, Shorts, live streaming, subscriptions, YouTube Music, creator monetization, recommendations, and ads.

You are investigating a reported decline in advertiser retention on YouTube. The issue affects advertisers who buy media across YouTube surfaces and ad formats, potentially including self-serve SMBs, agencies, large brands, performance marketers, and app-install advertisers. The decline may reflect advertisers stopping spend, reducing campaign activity, failing to renew, or shifting budget to competing channels such as Meta, TikTok, Amazon, or other digital ad platforms.

Your task is to lay out an analysis plan for diagnosing the root cause, not to propose a final fix upfront. The plan should clarify how retention is defined, confirm whether the decline is real, identify which advertiser segments or workflows are affected, and connect the metric movement to possible changes in auction dynamics, campaign performance, product experience, policy, pricing, measurement, seasonality, or external market conditions.

The investigation should account for YouTube’s scale and complexity: advertisers interact through Google Ads, campaign setup flows, targeting and bidding systems, creative requirements, reporting dashboards, billing, account support, and privacy-preserving measurement. A strong RCA plan should separate data issues from actual behavioral change, then progressively narrow the problem through segmentation, funnel analysis, and evidence-backed hypotheses.

The experience should consider:

- The exact retention metric definition, including advertiser denominator, time window, spend threshold, renewal/repeat-spend criteria, and whether retention is measured by count of advertisers, spend, campaigns, or active accounts.

- Instrumentation and data-quality checks to validate whether tracking, account identity, billing, campaign status, or reporting changes could be creating an artificial decline.

- Segmentation by advertiser size, geography, vertical, acquisition cohort, tenure, objective, campaign type, ad format, device, bidding strategy, and managed versus self-serve accounts.

- Funnel and lifecycle analysis across onboarding, campaign creation, approval, delivery, optimization, reporting, billing, and repeat-budget allocation.

- Hypotheses around advertiser value, including ROI/ROAS changes, reach, conversion measurement, targeting effectiveness, creative performance, auction costs, brand safety, policy friction, and support experience.

- External and ecosystem factors such as macro budget pressure, seasonality, competitor performance, privacy changes, regulatory shifts, or agency-level budget reallocations.

- Evidence needed to prioritize hypotheses, including time-series correlation, cohort comparisons, experiment logs, product launch calendars, support tickets, advertiser surveys, sales feedback, and competitive signals.

- Mitigation and prevention considerations, including how to monitor leading indicators, communicate with sales/support teams, and determine whether emergency action is required.

The goal is to present a structured, MECE analysis plan that helps YouTube quickly determine whether advertiser retention decline is a measurement artifact, a localized segment issue, or a broad deterioration in advertiser value—and to identify the highest-confidence path toward mitigation without jumping prematurely to a solution.