Pulling forward sparse data

Why would I need to pull forward sparse data?

• Many non-finance tools don’t report data over time, or leave out months without activity.
• A deal from a CRM might only have one ‘month’ it was created or closed in.
• A user sign-up in your product might only have a date associated with the first login

• In your model, you might care about these things over a longer period than the month the source systems report in.
• Just because a deal was created in January, doesn’t mean its ACV is zero in March;
• Just because a user signed up in April, doesn’t mean they are shouldn’t be counted as a user in June;

• This is especially true when you want to model based off of some ‘initial value’ over a longer period of time:
• spreading out a contract value into recognized revenue over the course of a year
• applying cohort growth/decay curves to historical data that started in a single month

OK, how do I do it?

• Integration data
• When configuring which drivers you want to bring into Runway, you’ll see an option on the right-hand side you can turn on if it’s a value you want to fill in months that may not have data in the source system.
• To see how the source reports data, you can use the data preview (1), or confirm the configuration, check your output, and tweak the configuration accordingly.
• With this option enabled, those drivers will automatically have the “Most Recent Value” time range option applied.
• You can also update the time range on specific formula pills related to integration data to pull forward the “Most Recent Value”, after you’ve stood up the database.

• General formulas
• When you have formulas that reference drivers, the same time range option for Most Recent Value is available in the filter menu options for time range.
• Note that “0” is considered as a valid value — the “Most Recent Value” logic only will pull forward in months that are coming back as NULL.

Examples