This article explains how you can use the platform to validate and measure your co-efficient factors.
INCRMNTAL developed methods to include coefficients to improve the accuracy of results by taking any external (or internal) factors into consideration.
What is Co-efficiency of Determination (R2)?
In statistics, the coefficient of determination, denoted R2 or r2 and pronounced "R squared", is the proportion of the variation in the dependent variable that is predictable from the independent variable(s).
It can be cold weather or sporting events that will raise the number of food deliveries, or a new year resolution that will increase the sales of gym memberships.
It can even be COVID lockdowns or releases of very popular movies that might impact your performance.
How can integrating coefficients to INCRMNTAL improve the results accuracy ?
Pointing out the external factors that impact your marketing performance means we can improve our prediction around those times, to take into consideration the coefficient factor.
Based on your historical data, in time the coefficient factor “kicks in”, we can train our prediction to expect the factor to impact your performance, and predict accordingly.
We can support training our model based on the following coefficient factors:
- Real world events - Weather, lockdowns, etc.
- Trending search terms - utilizing Google Trends (see note below)
- Stock tickers trends
- In-app events - Such as a limited offer, limited collections, etc.
- Supply and Demand
- External influencers - Such as iPhone launches or movie premieres.
- Holidays and weekends - Are already implemented in our platform.
** When basing co-efficiency on keywords, please mind those can’t be your brand keywords (i.e. your brand’s name), as this plays as the “chicken and the egg” - i.e. it is impossible to determine if those keywords are trending due to your marketing activities or vice versa.
From our experience, when it comes to knowing your co-efficiency factors and how they impact your performance, there are 3 stages you can be in:
I don’t know what my co-efficiency factors are
Let’s find out together - Explore measurements by manually adding activities, measuring, and removing if those are not influencers.
I’m pretty sure what my co-efficiency factors are
Let’s confirm that together - By providing activity data (excel, xml, csv, open source, list of keywords, etc…) as a one off to layer on top of our model, and re-measuring activities around said periods.
Applicable to customers with an Enterprise License Only
I know what my co-efficiency factors are
Great! We can train our model to know those co-efficiencies, by providing activity data (excel, xml, csv, open source, list of keywords, etc…) in an on-going way.
Applicable to customers with an Enterprise License Only
Interested in finding out what are your co-efficient factors? Contact us for further steps!
For any question please contact support@incrmntal.com , or open a support Ticket here