Skip to main content
The KPI is the business outcome your model tries to explain — the thing you want to understand what drove. Yapamar doesn’t limit you to a fixed list of metrics: any metric you upload becomes available for modeling. For example:
  • Sales revenue — how much money you made
  • Units sold — how many items you sold
  • Customer registrations — how many people signed up
  • New subscriptions, leads, bookings — or anything else your business measures
You upload your KPI data from a spreadsheet, and every numeric column in your KPI sheet is automatically detected as a metric. When creating a model, you pick the metric you want to model, and Yapamar figures out how much each marketing channel contributed to it.

​
How metrics are detected

Each column in your KPI sheet (other than the date and category columns) becomes a metric, named after its column header — a column called “Customer Registrations” shows up as the metric Customer Registrations throughout Yapamar. Only metrics that actually contain data are offered when you create a model, so an empty column won’t clutter your options. You can add new metrics at any time by uploading a file with additional columns — existing data is untouched, and the new metric becomes available for future models.

​
Metric and unit

When creating a model, the KPI block asks for three things:
  • Source — which of your data sources supplies the KPI
  • Metric — which uploaded metric to model
  • Unit — whether the metric is measured in money or in counts
The unit tells Yapamar how to treat and display your results. Monetary metrics are shown with your company’s currency (chosen when you registered). Count metrics are shown as plain numbers. The unit also drives how ROI is calculated: for monetary KPIs, ROI is the net return (incremental outcome − media investment) / media investment; for count KPIs it’s outcome per currency unit spent.

​
Combining metrics, models, and channels

You can build multiple models on the same uploaded KPI data. Any metric × channel selection is a valid model configuration:
  • Sales × Facebook + Google
  • Sales × Facebook + Google + TV
  • Units Sold × Facebook + Instagram
  • Registrations × Google + LinkedIn
Each combination is a distinct model with its own version history.