How investigate a brand-ad spend drop?

Meta offers a video ads product with two ad formats:

• Direct ads: tuned for actions that happen inside the platform
• Brand ads: when users click the video ad, they are taken to the advertiser's external site

An advertiser reports that its Brand Ads spend has fallen substantially during the last two weeks. As a data scientist, describe how you would:

1. Check whether the spend drop is real rather than an artifact of logging, reporting, or seasonality.
2. Assess whether the issue is limited to this advertiser or reflects a broader product or system change.
3. Identify the most probable root cause.
4. Propose follow-up analyses or experiments.

In your response, specify the key metrics you would examine, show how you would break down spend, discuss possible confounders, and explain how you would keep placement mix, geography mix, or attribution differences from producing misleading conclusions.

Overview: This question tests a data scientist's ability to diagnose ad-spend anomalies, covering data validation, attribution and reporting integrity, segmentation and placement effects, and causal inference in advertising analytics.