---
title: "Creating a new Multi Fit model"
slug: "creating-a-new-multi-fit-model"
updated: 2025-08-01T13:12:46Z
published: 2025-08-01T13:12:46Z
canonical: "help.madkudu.com/creating-a-new-multi-fit-model"
---

> ## Documentation Index
> Fetch the complete documentation index at: https://help.madkudu.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Creating a new Multi Fit model

This comprehensive guide provides a 3-week approach to creating a new multi-fit model from initial assessment through deployment. The process balances data-driven insights with business intelligence, ensuring your new model not only performs statistically well but also supports your sales team's daily operations and strategic objectives.

The methodology outlined here emphasizes validation at every step, cross-functional collaboration, and iterative improvement to deliver a model that drives measurable business impact.

## Prerequisites - Week 1

- Confirm which integrations are connected
- Confirm which audience and conversion mappings should be used to measure the performance of the model
- [Assess if a new multi-fit model is necessary by checking the performance of the current fit models against this audience and conversion](/v1/docs/i-have-a-new-segment-do-i-need-a-new-model)
- Internally align on the new ICP of the new model
- Internally align on the business rules the new model should follow (example: always qualify Fortune 500 companies)
- Confirm that the fields to receive new scores have been created in the destination CRM

## Step by step configuration - Week 2

- [Update audience](/v1/docs/audience-mapping) and [conversion mappings](/v1/docs/conversion-mapping), if necessary
- Create a new fit model
- [Create dataset with the relevant audience and conversion](/v1/docs/customer-fit-training-and-validation-datasets)
- [Create the trees](/v1/docs/customer-fit-how-to-create-or-edit-a-decision-tree) based on

-The new target ICP

-[The Studio Insights statistics](/v1/docs/customer-fit-insights)

- Adjust the thresholds
- [Add overrides based on business rules](/v1/docs/customer-fit-overrides)
- [Configure the signals](/v1/docs/customer-fit-signals)
- Review the [performance](/v1/docs/evaluate-your-draft-customer-fit-performance-against-your-live-model) of the model and a validation sample

## Feedback and iteration - Week 2&3

- Assess which persona(s) fall under which fit segment, and validate internally
- Do several sample reviews with peers as a sanity check
- Share some examples with Sales to get their feedback
- Iterate on model configuration based on feedback

## Go Live - Week 3

- Deploy the new model
  - Consider deploying on a Friday afternoon to avoid disrupting business

- Edit the Push configuration to push new scores to new fields - **Has to be done by a MadKudu Support team member**

**Estimated time: 15 hours**