Approach for customer segmentation

Survey

Post-purchase follow-up

Net Promoter Score

Customer Satisfaction Score

Collect data
about your existing and potential customers.

From phase 1 to phase 2

Define the characteristics of customers who are a fit for your products.

From phase 2 to phase 3

Roll out your marketing analytics and tactics —
collect more data.

From phase 3 to phase 4

Set predictive algorithms and probabilities for each individual or business

From phase 4 to phase 5

GOALS

PHASE 5. Predictive Analysis

PHASE 4. Run Analytics

PHASE 3. Implement and manage data

PHASE 2. User Segmentation model

PHASE 1. Business Analysis

Parachute USM - Approach for customer segmentation

Joint analysis of CRM data

Analysis of web performance data

Performance analysis on P1

Who are the policy buyers

Who are the ones who didn't buy

Is it possible to contact non-buyers to find out more about their intention (why they did not buy)

Is there a link between the purchases and the different communications sent

Sub-segment analysis

Data visualization dashboard

Complete report on relevant analysis of P1

Creating ease of use

Enough control for the user

Guidance

UX Challenges

Market segmentation

Geographic segmentation

Demographic segmentation

Cultural segmentation

Geographic & Demographic Segmentation

Behavioral segmentation

Segmenting customers by usage

Creating a clearer picture of customers

Behavioral and usage segmentation

How consumers see themselves and their world ?

Researching psychographic profiles

Developing profiles of your customers and targets

Psychographic segmentation

Why do people buy your product?

Opportunities to reach new markets

Benefits-sought segmentation

Restructuring of the analysis system in place to serve the objectives of finalizing the USM

Implementation of new data acquisition method

Management of new knowledge and adaptation
to finalize phase 2

The objectives of phase 1 is to highlight the complete performance on phase 1 of the parachute project (P1). The idea is to use the data from P1 to understand what worked well and led to policy purchases. Therefore, we will understand why a large majority of users did not purchase a parachute policy.

It will be then possible to create improvement hypotheses to be tested later on; in the next phases.

Thanks to the data structured in phase 1, we can now create new "User Segment Models".

In the past, demographic segmentation has been used to adapt content. In the Insurance business, we have noticed that most users expect detailed and customized user experiences. This is why phase 2 of the project will be used to create a new segment based on demographics data, but also behavioral and psychographic inputs: which will definitely boost the sales process regarding the different products.

If we want to validate models created in phase 2,
we need to change the way we acquire our data.

Phase 3 will include the creation of an analysis plan allowing us to adjust our data structures so as to confirm or refute our segmentation assumptions. We will also ensure that we have the most precise analytics along with the most relevant structure ; as to better understand our customers and to be able to adapt our sales tunnel — according to users reactions in response of various stimuli that we send them.

This phase will help us to provide feedback on our new models. It will ensure that the structures created will confirm quality of the anticipated data.
At this stage, we will launch a serie of tests and we will finally be able to have enough data to properly segment our users. It is also important in this phase to test certain additions in the processes which will allow us to better know our users (ex: qualification question).

This phase gives us control over all of our data sources. It is now possible to unify data to create a realistic portrait of our users. That way, this phase will allow us to define a predictive analysis structure based on segments previously tested in appropriate structures. Finally, we will replicate the models on new users . We should then be able to anticipate their potential to become a customer, abandonment rate, and the best structure to enable them to become Parachute customers.