# Measure a product launch across multiple social platforms

Keep platform metrics in their own defined columns and use a consistent business record for enquiries or orders. A person can encounter several channels, so adding platform reach or attributed outcomes does not automatically produce a unique audience or customer total.

Launch and actual offer:
Reporting window and time zone:
Platforms and published asset references:
Metric definitions per platform:
Paid activity per channel:
Campaign labels and destination routes:
Business enquiry or order definition:
Outcome source and anonymous reference method:
Cross-channel attribution rule:
Duplicate-outcome handling:
Unknown or overlapping sources:
Offer or stock changes during the window:
Channel observations:
Business outcome conclusion:
Next decision:


"record_type","channel","asset_reference","window","metric_name","value","unit_or_definition","paid_context","outcome_evidence","limitations"
"your observation","channel A","","","","","","","",""
"your observation","channel B","","","","","","","",""
"business outcome","combined business record","","","defined enquiries or orders","","apply the agreed duplicate rule","","",""


## Compare at the appropriate level

"Question","Record","Important boundary"
"What was published?","Actual asset references by channel","A draft is not a publication"
"What did each platform report?","Its own metric definitions and window","Same-looking labels can have different bases"
"What did the business receive?","Defined enquiry or order records","Apply one consistent outcome and duplicate rule"
"Which source is supported?","Actual attribution evidence","Retain unknown and overlapping sources"


## Example overlap

A fictional buyer might see a product on one channel, watch a demonstration on another and return directly to order. Three encounters are not three buyers. Your report should follow its actual attribution rule and keep the available evidence distinct from the story you infer.

Use the shared launch window to organise records, but preserve extraction time and observation age for individual assets. Record price, availability or distribution changes that affect the comparison.

## Before summing anything

- [ ] Units and definitions are compatible.

- [ ] Unique-person claims have an actual deduplication basis.

- [ ] Business outcomes follow one duplicate rule.

- [ ] Attribution gaps remain visible.

- [ ] Mixed paid and organic observations are labelled.