To achieve a robust model, it's crucial to think about and gather all the data that could impact the KPI you intend to measure.
The data you need includes:
Business KPI: The value Mortar MMM will forecast and optimise through its models (e.g. revenue, conversions, n. of leads, app downloads)
Marketing Campaigns: Paid Media activities with associated Spend (e.g. Facebook, Google) or Offline/other Paid activities with associated Spend (e.g. TV, radio, influencer)
Organic activities: Both Offline and Online marketing activities with no budget related (e.g. organic search impressions, email sent)
Contextual Variables: Organisation-level internal factors influencing sales (e.g. Promotions/discounts, price changes, new releases) or external factors beyond our control that may affect our business (e.g. Exchange Rates, COVID, Weather, Unemployment Rate, Market Trends)
Mandatory Data Requirements
Business KPI Data
You must include one business KPI (e.g. revenue, conversions, visits)
You can choose to include more but keep in mind you’ll need to build a separate model per each KPI you want to measure
Marketing Campaigns Spend Data
Spend data for all your marketing channels, including both digital and offline, are required
Optional Data Requirements
Engagement Metrics
Here are suggested metrics for various channels:
Digital activities: focus on impressions otherwise clicks
TV and Radio: focus on using Gross Rating Points (GRPs), Target Audience Rating Points (TARPs) or display impressions
Print (e.g. newspapers or magazines): include readership/reach metrics
Organic/non-paid Marketing Data:
Data such as social media impressions, newsletter emails sent/opened.
Contextual Data:
These refer to factors both internal and external to your business.
For example, they can be:
Promotions/discounts value (e.g. 15 for 15% discount)
Dates of new products released
Events like pandemics or weather