Is it literally possible to control for literally every single possible confounding variable?
No.
But, we don’t need to for this analysis to be powerful for you in your work. Here’s why:
Ultimately, the business you work with/for needs to make a decision on some aspect of their pricing strategy.
Implementing this type of approach gives them (and you) the best opportunity to make a winning pricing decision.
Using Accrue, if you’re able to control all major confounding factors, things like:
- Customer attributes - industry, segment, geography, etc
- Product attributes - term length, units sold, etc
- Sales team attributes - size of team, tenure, quota attainment
- Transactional attributes - number of deals, time of year, seasonality, etc
Is it possible that despite controlling for all those variables, there are other hidden variables at play?
Of course!
But is your estimate of impact reasonable, statistically sound, and defensible?
Does it allow a business to make the highest probability decision they can?
Yes, and yes.
In addition, as you identify hidden variables over time, you can add those variables to our model to train - continuing to strengthen your recommendations and predictions.
Sign up for a free pilot to explore further!
