Moost Assurance

Moost
by STEPS
Insurance edition | Insurers

Customer intelligence for insurers, beyond CRM

The software that tells you which policyholders are profitable, which are likely to leave and what to offer them, installed at the insurer.

BH Assurance proof of concept
Loss ratio per policyholder
Moost Assurance | Dashboard
Moost Assurance : dashboard

Application screenshot, fictitious data

The product

What is Moost Assurance?

Customer intelligence software, installed at the insurer. Connected to your existing data, it tells you which policyholders are profitable, which are likely to leave, what to offer them, and what each action brings in.

What a policyholder is worth
Loss ratio= claims ÷ premiums

What a policyholder costs in claims, compared with what they pay in premiums. The lower the loss ratio, the more profitable the policyholder. Moost calculates it for every policyholder, over 3 years.

600 claims / 1,000 premiums = 60%
Highly profitable
Profitable
Unprofitable
To review
0%30%70%150%
What Moost can bring

Each point of profitability is worth 1% of your premiums

Annual premiumsTND 100M
×
One point gained1%
=
Per yearTND 1M

One point less of loss ratio, or one point more of retention. Moost shows where to gain these points, policyholder by policyholder, and measures the gain against a control group.

Profitability and risk

Spot profitable policyholders about to leave

Every policyholder is placed by profitability and departure risk. Each cell has its recommended action, editable by your teams.

Low riskMedium riskHigh riskHighly profitable
Retain
Offer a 2nd policy
Priority agent call13,800 policies
Profitable
Nurture
Multi-policy
Reminder at D-60
Unprofitable
Monitor
Prevention
Review at renewal
To review
No reminder
Pricing review
Pricing review

Example: 13,800 highly profitable policies at high risk, the agents' priority.

The matrix in Moost Assurance
Renewal

Act before renewal, not after

The 10 playbooks follow the life of the policy. Renewal is at their heart.

D-90

Departure score

The non-renewal risk is calculated for every policy.

D-60

Agent call

Profitable policyholders at risk enter the call queue.

D-30

Final reminder

Reminder under a contact cap, through the preferred channel.

Renewal date

Measurement

Renewal rate compared with the control group.

Other playbooksSingle policyWin-backStated needUnpaid premium reminderPost-claim follow-upNew policy welcome
Insurance playbooks in Moost
360 profile

Every score is explained, policyholder by policyholder

Why this score? Policyholder P-00913
Request for a claims history+0.34
Premium increase of 5.4%+0.21
Renewal within 30 days+0.12
Customer for 7 years−0.08
No unpaid premium−0.05
Cancellation risk at D-9082%

In red, what raises the risk; in green, what lowers it. The agent knows what to say before calling.

The policyholder profile
Tour

Moost Assurance, screen by screen

Click a screenshot to enlarge it.

Screenshot, fictitious data
Dashboard

The portfolio at a glance

Active policies, 3-year loss ratio, retention and multi-policy, month on month.

  • Policyholders by profitability segment
  • Lifecycle stage
  • Actions awaiting approval
Screenshot, fictitious data
Predictions

Reliable scores, under control

Non-renewal at D-90 and multi-policy propensity at 180 days.

  • Minimum quality of AUC 0.65 before going live
  • Retrained every month, approved by a person
Screenshot, fictitious data
Targeting

Describe your audience in plain language

"Policyholders in the South zone driving an SUV, renewal in 30 days or less": 1,418 policyholders found, ready for an action.

  • Days to renewal, vehicle type, risk zone
  • Editable filters, export or action in one click
Screenshot, fictitious data
Actions

Safe actions, approved by your teams

Before anything is sent: exclusions (including policies cancelled for non-payment), consent, contact cap, control group and human approval.

  • The number of policyholders removed at each step is displayed
  • Sent by file, signed webhook or API
Screenshot, fictitious data
Measurement

The gain is proven, in premiums saved

Renewal at D-60: 81.2% among treated policyholders versus 72.4% in the control group, i.e. +8.8 points and TND 412,000 of additional revenue.

  • Return on investment shown as soon as the cost is entered
Screenshot, fictitious data
Data

Connected to your data, read-only

Six sources: policyholders, policies, premium notices, claims, renewals, agents. Through files, Oracle, SQL Server, PostgreSQL, MySQL, MariaDB databases, SFTP or API.

  • One policyholder, one profile: matching by national ID, customer ID or name, city and date of birth
Comparison

CRM manages the relationship, Moost drives profitability

Moost does not replace the CRM: it feeds it the right policyholders, at the right time, with the reason for the action.

A CRM

"Who is this customer, and what happened with them?"

Moost Assurance

"Which policyholders to act on, why, and what does it bring in?"

CRMDynamics 365, Salesforce, HubSpotMicrosoft Customer InsightsCustomer data and journeysMoost AssuranceInsurance customer intelligence
SegmentationPARTIAL
Lists and filtered views
YES
Rule-based segments, with Copilot
YES
Insurance segments, plain-language search
Loss ratio per policyholderNO
Premiums and claims outside the CRM
NO
Indicator to model
YES
Over 3 years, 4 segments
Cancellation riskNO
Lead and opportunity scores
PARTIAL
Generic churn
YES
At D-90, explained per policyholder
Multi-policyNO
No model
PARTIAL
Recommendation from purchases
YES
Propensity at 180 days
Actions and campaignsYES
Tasks and campaigns
YES
Multichannel journeys
YES
10 insurance playbooks
Proof of gainPARTIAL
Clicks and conversions
PARTIAL
Journey engagement
YES
Control group, premiums saved
Data and AI at the insurerPARTIAL
Cloud; Dynamics 365 on-premises until 2029
NO
Azure cloud, Copilot in the cloud
YES
Your servers, local AI
Security and deployment

Installed on your premises, built for sensitive data

On your servers

No data leaves your network. Installation possible without internet access.

Protected personal data

Identifiers replaced by a hash, personal data encrypted, export and erasure on request.

Controlled access

One-time code or company directory, four roles.

Local AI

Models and generative AI on your servers, with no personal data transmitted.

Traceability

Chained, verifiable audit log of every action.

Operations

Tested backups, updates with automatic rollback.

Getting started

Start with a pilot line of business

Step 01

Data workshop

Inventory of sources and choice of the pilot line, for example motor.

Step 02

Installation

On your servers, with 3 years of history and the loss ratio of every policyholder.

Step 03

Scores and playbooks

Models trained on your data, two or three playbooks launched.

Step 04

Measurement and decision

Premiums saved against the control group, then the decision to scale.

Commercial model

Perpetual licence for the Assurance edition and an annual maintenance contract. Proposal based on your volumes and scope.

Proof

Born from a project with BH Assurance

Moost Assurance comes from a proof of concept run with BH Assurance, in motor insurance.

FAQ

Frequently asked questions

1. Does Moost Assurance replace our CRM or core insurance system?

No. Moost reads your policies, premium notices and claims read-only and sends agents and the CRM the right policyholders, with the reason for the action.

2. What data do you need?

Ideally 3 years of history: policyholders, policies, premium notices, claims, renewals and agents, through files, database or API.

3. Does our data leave the company?

No. Moost Assurance is installed on your servers, the AI runs locally and installation is possible without internet access.

4. Can agents understand the scores?

Yes. Every score is explained by its main factors, in plain words, on the policyholder profile. The agent knows what to say before calling.

5. How do we get started?

With a pilot on one line of business, for example motor: data workshop, installation, two or three playbooks, then a decision on premiums saved.