Results in practice

Results in practice

Four cases from practice - at four leading brokers in the Netherlands. Concrete figures, real conversations, measurable impact.

Market context
~80%
of insurers run AI pilots
7%
successfully scales to production

The gap between experimentation and scale is wide. Our clients are in that 7% — the cases below show how.

Source: BCG, Build for the Future 2024 Global Study

The four cases at a glance
Why these cases are anonymised

Our clients use Onesurance to build a data lead in their market - and that is strategic.

That is why we do not publicly share names and exact results. In a personal conversation, Onesurance - with consent - shares the client name behind the case and the relevant contacts.

Data processed within client environment
GDPR-compliant · ISO 27001-certified infrastructure
Audit trail · duty of care demonstrable
Human-in-the-loop: advisor decides and conducts the conversation
1

Case study

Advisory chain · part of major Dutch insurer

Client since May 2024

How does an advisory chain stop lapse — before clients cancel?

Starting point
Lapse rate 14% · book-wide
Scope
All offices · ~85 advisors
Approach
Top Defend + Sales + AI Agent
Result
−75% lapse · €950K CLV

Advisory chain within major Dutch insurer. ~85 advisors, multiple offices, personal and commercial lines.

The approach

May – Jun 2024

EDA & business case

Book, data quality and lapse patterns analysed. Baseline: 14% lapse. Model: 10× more accurate than random selection.

Nov 2024 – Feb 2025

Pilot Top Defend

Lapse predictions live at 5 offices. Weekly feedback cycles with advisors. First retention results visible.

Mar 2025 – present

Rollout Top Defend + Sales + AI Agent

Live across all ~85 advisors. Contact rate from 50% to 75%. AI Agent active at pilot offices for outreach via email and chat.

"We often only called clients after they had already cancelled. It felt like turning up late to the party."

Project manager · advisory chain

"AI showed us precisely which clients needed additional cover. Hit ratio of more than 90%. The AI predicts it - the advisor makes it happen."

Project manager · advisory chain
Pilot → production in 11 months
Results · A/B test control group · 5,829 leads
−75%
lapse in contacted clients
12.9% → 3.3% personal · 11.9% → 4.1% commercial
€950K
customer value retained
CLV contacted group
€125K
commission loss prevented · Jan 2025 – Dec 2025
Modules deployed
Top DefendLapse risk at client and policy level
Top SalesCross- and upsell based on coverage ratio and client profile
AI AgentPersonalised outreach via email and chat
Curious what Onesurance will find in your book? We will show you live, using your own data.
Book a meeting →
2

Case study

National advisory chain · part of insurance holding

Client since March 2024

How does a large consolidator actively steer retention — across all brands at once?

Starting point
Lapse rate ~9% · silent attrition
Scope
~70K relationships · ~230K policies
Approach
Top Defend + Sales + Care + Fit
Result
−60% lapse · €400K CLV

National advisory chain, part of insurance holding. ~70K relationships, ~230K policies, multiple brands.

The approach

Mar – Apr 2024

EDA & business case

Data quality, lapse, CLV and cross-sell analysed across the full book. Baseline: ~9% lapse. Top 1% risk 5× more accurate than random.

Oct 2024 – Jan 2025

Pilot Top Defend

Monthly work lists with at-risk clients across multiple offices. Advisors logged follow-up; model learned every round.

Feb 2025 – present

Rollout — one work list for all modules

Top Sales, Top Care and Top Fit added. All advisors in one portal with management dashboard. Parent holding approves rollout to remaining brands.

"In clients we actually spoke to, lapse rates dropped by around 60%. The data consistently identified the right client - all we had to do was have the conversation."

Onesurance Quarterly Business Review · February 2026

"No more loose Excel exports. One work list, one dashboard, clear choices. Advisors use it every day."

Business owner · national advisory chain
The conclusion

Nearly €400,000 in retained customer value across more than 1,000 targeted conversations - with a payback period of less than one month.

Leadership is targeting an EBITDA uplift of around 30% once the approach is fully embedded. Not by hiring more people, but by enabling advisors to pick sharper - who to talk to, when, and about what. Next phase: rollout to the remaining brands within the holding and an AI agent that partially automates portfolio review conversations.

Pilot → production in 9 months
Results · measured vs. baseline · ~1,000 leads
−60%
lapse in contacted clients
13.2% → 5.5%
€400K
customer value retained
CLV contacted group
€70K
additional commission realised · 2025 – Q1 2026
Modules deployed
Top DefendLapse risk at client and policy level
Top SalesCross- and upsell based on coverage ratio and client profile
Top CareDuty of care index with audit trail
Top FitProfitability analysis at relationship level
Running a holding with multiple brands too? We will show you how one approach pays off across brands.
Book a meeting →
3

Case study

Top-10 broker · Netherlands · MGA

Client since May 2025

More value per advisor — without adding capacity

Starting point
Rising lapse · untapped sales potential
Scope
~120K personal · ~25K commercial
Approach
Top Defend + Sales + Care + agent
Result
−90% commission loss · +€1M/year

Top-10 broker with MGA. ~120K personal and ~25K commercial relationships, multiple offices.

The approach

Mar – Apr 2025

EDA & business case

Business Impact Assessment across the full book. Success criteria and pilot offices established. Model: 5.9× more accurate than random selection.

May – Aug 2025

Pilot Top Defend + Sales

Biweekly work lists at pilot offices. Seven rounds with stand-ups and monthly management reviews. A/B test across ~2,500 relationships confirms impact.

Nov 2025 – present

Rollout — all offices + Top Care

Live at all offices. Top Care added. Adoption ~70%, top office 95%. AI Agent for automatic scheduling in preparation (Q2 2026).

"For contacted relationships we saw commission losses drop by roughly 90%. The data spotted it before the advisor did."

Onesurance QBR report · February 2026

"The most enthusiastic office shows what is possible: 95% adoption, 67% contact rate - and the only office with a positive commission delta."

Internal analysis · top-10 broker
● Top Defend Clients retained through early signals
✓ Client retained
Review meeting scheduled - full package retained
Onesurance flags lapse risk. Advisor schedules a review meeting. Without this signal the client would have left quietly.
"Review meeting scheduled for 12-03-2026 at our office."
~€300/yr
commission retained
✓ Full package
no policies lost
✓ +6 policies
Defend contact leads to 6 additional policies
Advisor calls regarding a lapse signal. During the conversation several gaps in cover come to light - caravan, travel, legal expenses and motor.
"Email sent to review the package. Caravan, travel, legal expenses and motor as cross-sell."
+6 policies
via 1 Defend contact
+~€450/yr
additional commission
● Top Sales Growth opportunities converted into new policies
✓ Quote scheduled
Client buying property - full package quote
Onesurance detects purchase behaviour. Advisor calls - the client is indeed buying a property. Full package discussed.
"Buying a property, quote full package and discuss 21-01-2026."
+5 policies
potential · 4 lines
✓ Quote confirmed
✓ Appointment same day
Client calls back immediately - open to extending cover
Lead created by the system. Advisor calls. The client calls back the same day and is open to additional insurance.
"Client called back immediately. Open to additional insurance."
+3 policies
potential · 2 lines
✓ Same day
called back
✗ Missed opportunity What happens when leads are not followed up

The advisor chose not to make contact - an active decision. The data shows that contacted relationships cut commission losses by around 90% compared to non-contacted relationships.

✗ Client lost
Over €40,000 CLV lost - "no contact needed"
Onesurance places relationship in Top Defend. Advisor: "no contact needed." Client moves to a competitor. The AI had correctly assessed the risk.
"Client is managed by a competitor [...] placed elsewhere."
−~€9,500/yr
commission lost
−~€40,000
CLV lost
✗ Client lost
Over €24,000 CLV lost - earlier contact, still left
After earlier contact moments the advisor decides that no further follow-up is needed. Onesurance had flagged the relationship again as a risk. The client leaves regardless.
"Contact made several times mid-2025. No contact needed."
−~€7,500/yr
commission lost
−~€24,000
CLV lost
The conclusion

Over €1 million in additional book value per year - with advisor capacity unchanged.

The difference is not in working harder, but in knowing which relationship deserves attention and when. The AI flags. The advisor makes it happen. Adoption now sits at around 70% and is still climbing - the next phase, automatic scheduling via an AI agent, puts 90%+ within reach.

Pilot → production in 6 months
A/B test · contacted vs. not contacted · ~2,500 relationships
−90%
commission loss in contacted clients
contacted vs. not contacted
+€1M
additional portfolio value per year
with unchanged advisor capacity
95%
adoption · top office · Nov 2025 – Apr 2026
Modules deployed
Top DefendLapse risk at client and policy level
Top SalesCross- and upsell based on coverage ratio and client profile
Top CareDuty of care index with audit trail
AI AgentAutomatic scheduling (Q2 2026)
Want demonstrable impact per advisor too? We will run a similar A/B test on your own book.
Book a meeting →
4

Case study

Independent advisory group · multiple offices · personal & commercial · part of German consolidator

Client since October 2025

From Excel lists to a data-driven client rhythm — in one quarter

Starting point
Lapse rate ~14% · Excel work lists
Scope
~65K relationships · ~250K policies
Approach
Top Defend + Sales · all advisors
Result
0 lapses · ~50 additional policies

Independent advisory group with own MGA. ~65K relationships, ~250K policies, personal and commercial.

The approach

Oct – Nov 2025

EDA & business case

Data and business check on the full book. AI engine trained on own data. Baseline: ~14% lapse. Model: 100% precision on top list in the first quarter.

Dec 2025 – Mar 2026

Pilot Top Defend + Top Sales

Biweekly work lists at all offices. Five rounds with bidirectional feedback between management and advisors.

Apr 2026 – present

Rollout — Top Care, Acquire & agents

Top Care and Top Risk added. Acquire pilot for commercial prospects. Defend mail + chat in preparation. New advisor portal H2 2026.

"Every two weeks we get exactly the relationships we should have been on top of in advance. Advisors no longer search - they call."

Business owner · independent advisory group

"On the top risk list we have not seen a single cancellation this quarter. That says everything about the timing of the signals."

Onesurance Quarterly Business Review · Q1 2026
The conclusion

A working rhythm of data-driven client management across all offices - within one quarter.

No longer a pilot tucked away in a corner, but part of every advisor's biweekly cadence. The next phase - Acquire and the first Defend agents - brings new acquisition and automation within reach.

Pilot → production in 3 months
Results · pilot Q1 2026 · ~450 leads
0
lapses on the Defend top list
Jan – Apr 2026
>€120K
customer value retained
CLV · contacted Defend group
~50
additional policies from Top Sales · ~25 lines extended
Modules deployed
Top DefendLapse risk at client and policy level
Top SalesCross- and upsell based on coverage ratio and client profile
AcquireCommercial prospect identification (Q2 2026)
Defend AgentPersonalised outreach via email and chat (Q2 2026)
Quick start possible? Yes - from data connection to production in a single quarter is realistic.
Book a meeting →

Curious what this could deliver for your firm?

Book a 30-minute conversation. We will show you live what Onesurance sees in your book.

Book a demo → View the platform