Diagnose a sudden drop in seller liquidity for inventory forecasting panel

Problem Statement Description

You are investigating a sudden drop in seller liquidity among freelancers who use an inventory forecasting panel. In this context, seller liquidity refers to the availability and responsiveness of freelancer supply: whether sellers have enough forecasted inventory, capacity, or listed availability to fulfill buyer demand without excessive delays, stockouts, cancellations, or stale listings.

The drop appears abrupt rather than gradual, so the interview should focus on diagnosing whether this is a real marketplace supply issue, a measurement or instrumentation problem, a product workflow regression, or a segment-specific behavior change among freelancers. The inventory forecasting panel is a key surface because freelancers may rely on it to decide how much inventory or capacity to list, when to replenish, how to price, and whether to accept demand.

Before recommending fixes, frame the anomaly clearly, isolate where the drop is happening, and evaluate competing hypotheses using data, user workflow evidence, and recent system or business changes.

The experience should consider:

- How “seller liquidity” is defined, including the numerator, denominator, time window, and whether it measures active sellers, available inventory, fulfillment readiness, response rates, or successful matches.

- Whether the drop is isolated to freelancers or also appears across agencies, enterprise sellers, regions, categories, tenure cohorts, devices, acquisition channels, or pricing tiers.

- Instrumentation checks for tracking changes, event loss, schema changes, delayed pipelines, dashboard logic, bot filtering, deduplication, or changes in how inventory availability is counted.

- Recent product, policy, algorithmic, pricing, notification, forecasting-model, or onboarding changes that could have affected how freelancers interpret or act on the panel.

- Funnel points in the seller workflow, such as opening the panel, viewing forecasts, editing inventory, saving updates, publishing availability, accepting orders, and replenishing supply.

- External and marketplace factors, including demand spikes, seasonality, freelancer churn, payment delays, trust and safety actions, supply constraints, or category-specific disruptions.

- Evidence needed to distinguish correlation from causation, such as before/after comparisons, unaffected control cohorts, logs, session replays, support tickets, seller interviews, and operational alerts.

- Immediate mitigation and prevention considerations, including how to stabilize the marketplace while maintaining buyer trust, seller confidence, and data integrity.

The goal is to demonstrate a structured root-cause investigation: define the anomaly, validate the data, segment the impact, generate and test hypotheses, identify the most likely cause or causes, and outline what evidence would be required before moving into fixes.