You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 3 Next »

The pricing recommendation tool aims at generating reliable price recommendations based on AI identification of most comparable cpcs :

1.Collect price drivers

  • Collected a list of potential price drivers

  • Built Data pipeline and feature engineering

  • Validate shortlist with product managers

2. Identify most similar cpc's

We have Built Machine Learning model to identify nearest neighbor CPCs based on price drivers
Quantifying comparability of CPCs using a 'similarity score' and Identify most comparable ones
After that we apply adjustments & hard boundaries to maximize reliability, based on business rules.

3. Generate Price recommendations

To select final comparable set we define a threshold of similarity and we built a pricing mechanism to generate the price recommendation out of the comparable set.
We make an adjustment of top comparable CPC prices to account for potential discounts effect on larger volumes, and
we compute final recommendation through the median of comparable CPCs price points

4. Incorporate Feedback Mechanism

Account Managers are able to modulate recommendations by:

  • Incorporate business overlays (i.e. to cap a price increase at a customer level based on historical data).

  • Deselection of a comparable that seem not pertinent.

  • No labels