How to Use Databox for Source-system number checks

Databox works well for source-system number checks when you run it like production work: locked brief, SOURCE facts, then review before publish. Databox is AI-powered business analytics that answers performance questions from connected data (databox.com). Catalog notes a free tier and custom plans; verify databox.com/pricing. Validate numbers against source systems before sharing. Start at /explore/databox.

This guide focuses on source-system number checks in detail. Related Databox articles: /blog/how-to-use-databox-for-slack-or-email-snapshots, /blog/how-to-use-databox-for-metric-definition-reviews, /blog/how-to-use-databox-for-board-ready-report-gates.

When this workflow is the right job

Use source-system number checks when the deliverable is specifically this Databox job. Switch to metrics source connections when that workflow already owns the asset.

Step by step workflow

1. Brief Source-system number checks

Write what must stay true for source-system number checks in Databox before settings or spend.

Brief: Source-system number checks
Keep: verified SOURCE facts only
Avoid: invented pricing or features
Success: one reviewable output

2. Open Databox for Source-system number checks

Use the Databox surface that owns source-system number checks. Do not mix a neighboring workflow in the same pass.

Surface: Source-system number checks
Start: pilot with one representative input
Plans: databox.com/pricing

3. Pilot Source-system number checks

Run a single source-system number checks pilot. Score clarity, grounding, and whether the output is reviewable.

Pilot: Source-system number checks
[ ] SOURCE facts match
[ ] Output reviewable
[ ] Settings logged

4. Refine Source-system number checks

Change one source-system number checks dimension only. Save a template from the best run.

Refine: Source-system number checks
Change: one control only
Keep: SOURCE and success criteria

Practical source-system number checks examples

Embedded deploy note

Scenario:
An agency analyst runs Source-system number checks in Databox for "Embedded deploy note".

Objective:
Deliver an executive-ready analysis for Embedded deploy note with defined metrics and date range.

Inputs:
- Client name
- Metrics and date range for Embedded deploy note
- Segment definitions
- Decision needed (e.g., budget shift)

Workflow:
Confirm connected data → Query Source-system number checks → Validate totals → Draft recommendations for Embedded deploy note → Client review

Requirements:
- Stay within verified Databox capabilities; do not invent features.
- Confirm live plan notes on databox.com/pricing before promising volume.
- Change one variable between iterations.
- Human-review before external publish, send, billing, or clinical/legal use.
- Do not invent metrics not present in connected data.

Expected output:
A summary for Embedded deploy note with metric table, caveats, and a recommended next action.

Data source connect

Scenario:
An agency analyst runs Source-system number checks in Databox for "Data source connect".

Objective:
Deliver an executive-ready analysis for Data source connect with defined metrics and date range.

Inputs:
- Client name
- Metrics and date range for Data source connect
- Segment definitions
- Decision needed (e.g., budget shift)

Workflow:
Confirm connected data → Query Source-system number checks → Validate totals → Draft recommendations for Data source connect → Client review

Requirements:
- Stay within verified Databox capabilities; do not invent features.
- Confirm live plan notes on databox.com/pricing before promising volume.
- Change one variable between iterations.
- Human-review before external publish, send, billing, or clinical/legal use.
- Do not invent metrics not present in connected data.

Expected output:
A summary for Data source connect with metric table, caveats, and a recommended next action.

Null metric handle

Scenario:
An agency analyst runs Source-system number checks in Databox for "Null metric handle".

Objective:
Deliver an executive-ready analysis for Null metric handle with defined metrics and date range.

Inputs:
- Client name
- Metrics and date range for Null metric handle
- Segment definitions
- Decision needed (e.g., budget shift)

Workflow:
Confirm connected data → Query Source-system number checks → Validate totals → Draft recommendations for Null metric handle → Client review

Requirements:
- Stay within verified Databox capabilities; do not invent features.
- Confirm live plan notes on databox.com/pricing before promising volume.
- Change one variable between iterations.
- Human-review before external publish, send, billing, or clinical/legal use.
- Do not invent metrics not present in connected data.

Expected output:
A summary for Null metric handle with metric table, caveats, and a recommended next action.

Date range lock

Scenario:
An agency analyst runs Source-system number checks in Databox for "Date range lock".

Objective:
Deliver an executive-ready analysis for Date range lock with defined metrics and date range.

Inputs:
- Client name
- Metrics and date range for Date range lock
- Segment definitions
- Decision needed (e.g., budget shift)

Workflow:
Confirm connected data → Query Source-system number checks → Validate totals → Draft recommendations for Date range lock → Client review

Requirements:
- Stay within verified Databox capabilities; do not invent features.
- Confirm live plan notes on databox.com/pricing before promising volume.
- Change one variable between iterations.
- Human-review before external publish, send, billing, or clinical/legal use.
- Do not invent metrics not present in connected data.

Expected output:
A summary for Date range lock with metric table, caveats, and a recommended next action.

Segment top 3

Scenario:
An agency analyst runs Source-system number checks in Databox for "Segment top 3".

Objective:
Deliver an executive-ready analysis for Segment top 3 with defined metrics and date range.

Inputs:
- Client name
- Metrics and date range for Segment top 3
- Segment definitions
- Decision needed (e.g., budget shift)

Workflow:
Confirm connected data → Query Source-system number checks → Validate totals → Draft recommendations for Segment top 3 → Client review

Requirements:
- Stay within verified Databox capabilities; do not invent features.
- Confirm live plan notes on databox.com/pricing before promising volume.
- Change one variable between iterations.
- Human-review before external publish, send, billing, or clinical/legal use.
- Do not invent metrics not present in connected data.

Expected output:
A summary for Segment top 3 with metric table, caveats, and a recommended next action.

Creative fatigue

Scenario:
An agency analyst runs Source-system number checks in Databox for "Creative fatigue".

Objective:
Deliver an executive-ready analysis for Creative fatigue with defined metrics and date range.

Inputs:
- Client name
- Metrics and date range for Creative fatigue
- Segment definitions
- Decision needed (e.g., budget shift)

Workflow:
Confirm connected data → Query Source-system number checks → Validate totals → Draft recommendations for Creative fatigue → Client review

Requirements:
- Stay within verified Databox capabilities; do not invent features.
- Confirm live plan notes on databox.com/pricing before promising volume.
- Change one variable between iterations.
- Human-review before external publish, send, billing, or clinical/legal use.
- Do not invent metrics not present in connected data.

Expected output:
A summary for Creative fatigue with metric table, caveats, and a recommended next action.

Geo split

Scenario:
An agency analyst runs Source-system number checks in Databox for "Geo split".

Objective:
Deliver an executive-ready analysis for Geo split with defined metrics and date range.

Inputs:
- Client name
- Metrics and date range for Geo split
- Segment definitions
- Decision needed (e.g., budget shift)

Workflow:
Confirm connected data → Query Source-system number checks → Validate totals → Draft recommendations for Geo split → Client review

Requirements:
- Stay within verified Databox capabilities; do not invent features.
- Confirm live plan notes on databox.com/pricing before promising volume.
- Change one variable between iterations.
- Human-review before external publish, send, billing, or clinical/legal use.
- Do not invent metrics not present in connected data.

Expected output:
A summary for Geo split with metric table, caveats, and a recommended next action.

Device mix

Scenario:
An agency analyst runs Source-system number checks in Databox for "Device mix".

Objective:
Deliver an executive-ready analysis for Device mix with defined metrics and date range.

Inputs:
- Client name
- Metrics and date range for Device mix
- Segment definitions
- Decision needed (e.g., budget shift)

Workflow:
Confirm connected data → Query Source-system number checks → Validate totals → Draft recommendations for Device mix → Client review

Requirements:
- Stay within verified Databox capabilities; do not invent features.
- Confirm live plan notes on databox.com/pricing before promising volume.
- Change one variable between iterations.
- Human-review before external publish, send, billing, or clinical/legal use.
- Do not invent metrics not present in connected data.

Expected output:
A summary for Device mix with metric table, caveats, and a recommended next action.

Frequency cap

Scenario:
An agency analyst runs Source-system number checks in Databox for "Frequency cap".

Objective:
Deliver an executive-ready analysis for Frequency cap with defined metrics and date range.

Inputs:
- Client name
- Metrics and date range for Frequency cap
- Segment definitions
- Decision needed (e.g., budget shift)

Workflow:
Confirm connected data → Query Source-system number checks → Validate totals → Draft recommendations for Frequency cap → Client review

Requirements:
- Stay within verified Databox capabilities; do not invent features.
- Confirm live plan notes on databox.com/pricing before promising volume.
- Change one variable between iterations.
- Human-review before external publish, send, billing, or clinical/legal use.
- Do not invent metrics not present in connected data.

Expected output:
A summary for Frequency cap with metric table, caveats, and a recommended next action.

Holdout note

Scenario:
An agency analyst runs Source-system number checks in Databox for "Holdout note".

Objective:
Deliver an executive-ready analysis for Holdout note with defined metrics and date range.

Inputs:
- Client name
- Metrics and date range for Holdout note
- Segment definitions
- Decision needed (e.g., budget shift)

Workflow:
Confirm connected data → Query Source-system number checks → Validate totals → Draft recommendations for Holdout note → Client review

Requirements:
- Stay within verified Databox capabilities; do not invent features.
- Confirm live plan notes on databox.com/pricing before promising volume.
- Change one variable between iterations.
- Human-review before external publish, send, billing, or clinical/legal use.
- Do not invent metrics not present in connected data.

Expected output:
A summary for Holdout note with metric table, caveats, and a recommended next action.

Confidence flag

Scenario:
An agency analyst runs Source-system number checks in Databox for "Confidence flag".

Objective:
Deliver an executive-ready analysis for Confidence flag with defined metrics and date range.

Inputs:
- Client name
- Metrics and date range for Confidence flag
- Segment definitions
- Decision needed (e.g., budget shift)

Workflow:
Confirm connected data → Query Source-system number checks → Validate totals → Draft recommendations for Confidence flag → Client review

Requirements:
- Stay within verified Databox capabilities; do not invent features.
- Confirm live plan notes on databox.com/pricing before promising volume.
- Change one variable between iterations.
- Human-review before external publish, send, billing, or clinical/legal use.
- Do not invent metrics not present in connected data.

Expected output:
A summary for Confidence flag with metric table, caveats, and a recommended next action.

Chart-ready table

Scenario:
An agency analyst runs Source-system number checks in Databox for "Chart-ready table".

Objective:
Deliver an executive-ready analysis for Chart-ready table with defined metrics and date range.

Inputs:
- Client name
- Metrics and date range for Chart-ready table
- Segment definitions
- Decision needed (e.g., budget shift)

Workflow:
Confirm connected data → Query Source-system number checks → Validate totals → Draft recommendations for Chart-ready table → Client review

Requirements:
- Stay within verified Databox capabilities; do not invent features.
- Confirm live plan notes on databox.com/pricing before promising volume.
- Change one variable between iterations.
- Human-review before external publish, send, billing, or clinical/legal use.
- Do not invent metrics not present in connected data.

Expected output:
A summary for Chart-ready table with metric table, caveats, and a recommended next action.

Callout risks

Scenario:
An agency analyst runs Source-system number checks in Databox for "Callout risks".

Objective:
Deliver an executive-ready analysis for Callout risks with defined metrics and date range.

Inputs:
- Client name
- Metrics and date range for Callout risks
- Segment definitions
- Decision needed (e.g., budget shift)

Workflow:
Confirm connected data → Query Source-system number checks → Validate totals → Draft recommendations for Callout risks → Client review

Requirements:
- Stay within verified Databox capabilities; do not invent features.
- Confirm live plan notes on databox.com/pricing before promising volume.
- Change one variable between iterations.
- Human-review before external publish, send, billing, or clinical/legal use.
- Do not invent metrics not present in connected data.

Expected output:
A summary for Callout risks with metric table, caveats, and a recommended next action.

Next test idea

Scenario:
An agency analyst runs Source-system number checks in Databox for "Next test idea".

Objective:
Deliver an executive-ready analysis for Next test idea with defined metrics and date range.

Inputs:
- Client name
- Metrics and date range for Next test idea
- Segment definitions
- Decision needed (e.g., budget shift)

Workflow:
Confirm connected data → Query Source-system number checks → Validate totals → Draft recommendations for Next test idea → Client review

Requirements:
- Stay within verified Databox capabilities; do not invent features.
- Confirm live plan notes on databox.com/pricing before promising volume.
- Change one variable between iterations.
- Human-review before external publish, send, billing, or clinical/legal use.
- Do not invent metrics not present in connected data.

Expected output:
A summary for Next test idea with metric table, caveats, and a recommended next action.

Owner sign-off

Scenario:
An agency analyst runs Source-system number checks in Databox for "Owner sign-off".

Objective:
Deliver an executive-ready analysis for Owner sign-off with defined metrics and date range.

Inputs:
- Client name
- Metrics and date range for Owner sign-off
- Segment definitions
- Decision needed (e.g., budget shift)

Workflow:
Confirm connected data → Query Source-system number checks → Validate totals → Draft recommendations for Owner sign-off → Client review

Requirements:
- Stay within verified Databox capabilities; do not invent features.
- Confirm live plan notes on databox.com/pricing before promising volume.
- Change one variable between iterations.
- Human-review before external publish, send, billing, or clinical/legal use.
- Do not invent metrics not present in connected data.

Expected output:
A summary for Owner sign-off with metric table, caveats, and a recommended next action.

Q2 CPA ROAS

Scenario:
An agency analyst runs Source-system number checks in Databox for "Q2 CPA ROAS".

Objective:
Deliver an executive-ready analysis for Q2 CPA ROAS with defined metrics and date range.

Inputs:
- Client name
- Metrics and date range for Q2 CPA ROAS
- Segment definitions
- Decision needed (e.g., budget shift)

Workflow:
Confirm connected data → Query Source-system number checks → Validate totals → Draft recommendations for Q2 CPA ROAS → Client review

Requirements:
- Stay within verified Databox capabilities; do not invent features.
- Confirm live plan notes on databox.com/pricing before promising volume.
- Change one variable between iterations.
- Human-review before external publish, send, billing, or clinical/legal use.
- Do not invent metrics not present in connected data.

Expected output:
A summary for Q2 CPA ROAS with metric table, caveats, and a recommended next action.

Prospect vs retarget

Scenario:
An agency analyst runs Source-system number checks in Databox for "Prospect vs retarget".

Objective:
Deliver an executive-ready analysis for Prospect vs retarget with defined metrics and date range.

Inputs:
- Client name
- Metrics and date range for Prospect vs retarget
- Segment definitions
- Decision needed (e.g., budget shift)

Workflow:
Confirm connected data → Query Source-system number checks → Validate totals → Draft recommendations for Prospect vs retarget → Client review

Requirements:
- Stay within verified Databox capabilities; do not invent features.
- Confirm live plan notes on databox.com/pricing before promising volume.
- Change one variable between iterations.
- Human-review before external publish, send, billing, or clinical/legal use.
- Do not invent metrics not present in connected data.

Expected output:
A summary for Prospect vs retarget with metric table, caveats, and a recommended next action.

Budget +10% shift

Scenario:
An agency analyst runs Source-system number checks in Databox for "Budget +10% shift".

Objective:
Deliver an executive-ready analysis for Budget +10% shift with defined metrics and date range.

Inputs:
- Client name
- Metrics and date range for Budget +10% shift
- Segment definitions
- Decision needed (e.g., budget shift)

Workflow:
Confirm connected data → Query Source-system number checks → Validate totals → Draft recommendations for Budget +10% shift → Client review

Requirements:
- Stay within verified Databox capabilities; do not invent features.
- Confirm live plan notes on databox.com/pricing before promising volume.
- Change one variable between iterations.
- Human-review before external publish, send, billing, or clinical/legal use.
- Do not invent metrics not present in connected data.

Expected output:
A summary for Budget +10% shift with metric table, caveats, and a recommended next action.

Reach by channel

Scenario:
An agency analyst runs Source-system number checks in Databox for "Reach by channel".

Objective:
Deliver an executive-ready analysis for Reach by channel with defined metrics and date range.

Inputs:
- Client name
- Metrics and date range for Reach by channel
- Segment definitions
- Decision needed (e.g., budget shift)

Workflow:
Confirm connected data → Query Source-system number checks → Validate totals → Draft recommendations for Reach by channel → Client review

Requirements:
- Stay within verified Databox capabilities; do not invent features.
- Confirm live plan notes on databox.com/pricing before promising volume.
- Change one variable between iterations.
- Human-review before external publish, send, billing, or clinical/legal use.
- Do not invent metrics not present in connected data.

Expected output:
A summary for Reach by channel with metric table, caveats, and a recommended next action.

Client exec memo

Scenario:
An agency analyst runs Source-system number checks in Databox for "Client exec memo".

Objective:
Deliver an executive-ready analysis for Client exec memo with defined metrics and date range.

Inputs:
- Client name
- Metrics and date range for Client exec memo
- Segment definitions
- Decision needed (e.g., budget shift)

Workflow:
Confirm connected data → Query Source-system number checks → Validate totals → Draft recommendations for Client exec memo → Client review

Requirements:
- Stay within verified Databox capabilities; do not invent features.
- Confirm live plan notes on databox.com/pricing before promising volume.
- Change one variable between iterations.
- Human-review before external publish, send, billing, or clinical/legal use.
- Do not invent metrics not present in connected data.

Expected output:
A summary for Client exec memo with metric table, caveats, and a recommended next action.

Audience build time

Scenario:
An agency analyst runs Source-system number checks in Databox for "Audience build time".

Objective:
Deliver an executive-ready analysis for Audience build time with defined metrics and date range.

Inputs:
- Client name
- Metrics and date range for Audience build time
- Segment definitions
- Decision needed (e.g., budget shift)

Workflow:
Confirm connected data → Query Source-system number checks → Validate totals → Draft recommendations for Audience build time → Client review

Requirements:
- Stay within verified Databox capabilities; do not invent features.
- Confirm live plan notes on databox.com/pricing before promising volume.
- Change one variable between iterations.
- Human-review before external publish, send, billing, or clinical/legal use.
- Do not invent metrics not present in connected data.

Expected output:
A summary for Audience build time with metric table, caveats, and a recommended next action.

How to improve source-system number checks

Cut noise from source-system number checks by removing extra adjectives while preserving SOURCE facts in Databox.

Raise quality by insisting on a single success check before debating style.

Make review easier by labeling fields that must never change.

Speed iteration by cloning the last good run and altering only one control.

Stabilize outputs by pinning settings after the pilot is approved.

Reduce rework by rejecting drafts that invent claims.

Improve handoffs by recording which control produced the best result.

Harden the workflow by testing an incomplete input before trusting defaults.

Prompting and usage guidance

Name the source-system number checks job, audience, and success check before opening Databox.

Paste only verified facts under SOURCE so Databox cannot invent details.

Specify the deliverable shape up front.

Call out fixed details versus flexible style choices.

Ask Databox to flag unsupported claims before you accept the draft.

Limitations to respect

Check Databox plan gates for source-system number checks on databox.com/pricing before you promise timelines.

Keep drafts unpublished until a human confirms SOURCE facts.

Databox can be wrong. Treat source-system number checks as provisional until review.

If documentation is silent on a claim, leave it out rather than guessing.

Practical tips for this workflow

Pilot once before batching source-system number checks in Databox.

Keep a reusable template with variables for source-system number checks.

Separate creative instructions from SOURCE facts.

Log settings from the best run.

Common mistakes

• Skipping the pilot run before scaling volume
• Inventing pricing, quotas, or features not on official pages
• Mixing unrelated workflows in one session
• Publishing without a human review gate

Treat source-system number checks in Databox as a production workflow: brief, pilot, refine, then ship with review. Related reading: /blog/how-to-use-databox-for-slack-or-email-snapshots, /blog/how-to-use-databox-for-metric-definition-reviews, /blog/how-to-use-databox-for-board-ready-report-gates.