Skip to main content

What is Account Context in Adbrew?

Karan Saraf avatar
Written by Karan Saraf
Updated over 4 months ago

The Account Context feature allows you to provide background information that helps AI generate more accurate and meaningful summaries when reporting data on a dashboard. This ensures that AI considers relevant business events, strategies, or anomalies that raw data alone cannot explain.

What is Account Context?

Account Context lets you:

  • Add business context such as promotions, inventory issues, budget changes, or external factors.

  • Associate that context with specific date ranges and specific accounts.

  • Ensure AI-generated summaries (in text blocks or other widgets) are more relevant and accurate.

Example:

  • “Top three SKUs were out of stock due to shipment delays, causing sales to appear lower.”

  • “We increased ad spend significantly during Prime Day, boosting traffic and conversions.”


How to Add Account Context

1. Navigate to Settings > Account Context from the left sidebar menu.

2. Select the Date Range for which the context is valid.

3. Enter a Title for easy identification (e.g., Prime Day Strategy, Stock Shortage Impact).

4. Choose one or multiple Accounts to apply the context to.

5. Write the Context Description in plain text (business events, decisions, anomalies, etc.).

6. Click Save.

Context by Date Range & Account

  • You can assign context to specific date ranges.

  • You can assign context to specific accounts.

Best Practice:
Always separate context entries if they differ by account or date range.

  • Example:

    • Account A → “Stockouts impacted sales in July.”

    • Account B → “High budget allocated for August promotions.”


How Context Works with AI Summaries

  • When you generate AI summaries (e.g., via Text Block Widget or Dashboard summary button), the system will automatically pull in the Account Context relevant to the selected account and date range.

  • This ensures that summaries are tailored to real business situations instead of relying only on raw performance data.

Did this answer your question?