Libra Insurance doubled sales per agent in 90 days. A Sellence case study.

The company

A public-market insurer that already won product. The funnel leaked later.

Libra is Israel's first fully digital P&C insurer. Listed on the Tel Aviv Stock Exchange since 2021. Revenue grew from ₪300M in 2023 to ₪1.09B in 2025, with net income up 67% year over year. In the Capital Market Authority's 2024 Service Index, Libra ranked first in comprehensive motor insurance and third in home, a regulator-published, audited assessment, not a marketing claim.

The product was winning. The sales floor was working harder for less. Outbound calls reached fewer people every quarter. Agents spent the bulk of the day chasing rather than closing. Renewals stalled in the gap between an intent signal and the next available human.

The challenge

Phone-based selling was failing on a great product.

Every metric agents touched moved in the wrong direction at the same time. The team knew customers had moved to SMS and WhatsApp years ago. They needed to find out whether selling could follow them there, or whether the relationship could only survive on the phone.

Connect rate

Declining

Cold-call connect rates have fallen every quarter for three years. The pickup happens at a fraction of the rate it did pre-mobile.

Agent time

Misallocated

Reps spent most of their day chasing voicemail instead of closing deals already in motion.

Renewals

Stalling

Intent signals arrived faster than humans could respond. The cheapest deal to close was the renewal already in the pipeline. It was leaking.

The experiment

A real A/B test. Two teams. Same leads. One ran Sellence.

Libra didn't pilot. They ran a controlled experiment. Two teams of comparable headcount, same product, same pricing, same target segments, same lead source. One team kept the existing playbook. The other ran the same headcount with Sellence carrying the thread.

Control · No Sellence

Team A

Sales playbookPhone-led

Reply rateBaseline

EngagementBaseline

Starting positionAhead

Treatment · Sellence engine

Team B

Sales playbookSMS + WhatsApp, human hand-off

Reply rate+84%

Engagement+50%

Closed by day 90~2× Team A

Why this matters

The Sellence team started behind on close rate. Within 90 days they had nearly doubled the control. Two teams. Same product. Same pricing. Same leads. The only variable that moved was how the conversation was carried.

This isn't a pilot anecdote. It's a control vs treatment design with a publicly traded customer.

Closed deals per agent, day 0 to day 90

Indexed to Team A's day 0 performance. Team B (Sellence) starts behind on day 0 and crosses Team A inside three months.

Team A · Control Team B · Sellence

Note: Index reflects published 2× per-agent ratio at day 90. Daily curve is illustrative of trajectory; underlying weekly cohort data on file.

The mechanism

What the engine actually carried.

Sellence didn't replace agents. It carried the parts of the conversation that don't need a person, so the person could spend their hour on the parts that do.

01 · TIMING

The thread opens at the moment of intent.

Renewal window opens. Product page revisited. Quote abandoned mid-form. The first message lands within minutes, not next-day. The buyer is still in the decision.

02 · MEMORY

Every objection, kept on the record.

Every conversation, every preference, every quiet moment lives in the thread. When the agent steps in, they don't ask twice. The buyer never starts over.

03 · JUDGMENT

The right next move, surfaced for the rep.

Which question to ask. Which rebuttal already worked on this profile. Which deal is ready for a voice call and which one is closing on its own.

04 · HAND-OFF

A clean pass to the agent at the moment a voice is needed.

The engine carries the early conversation. The human carries the close. The buyer feels one continuous relationship, not a handoff between systems.

The playbook

Three things made the Sellence deployment decisive at Libra.

The result isn't an anecdote. It's a method. Three decisions did most of the work, and any agency selling a high-consideration B2C product can copy them.

The team got out of the chase.
- Sellence picked up inbound and warm-intent threads in minutes.
- Agents stopped chasing voicemail. They started the day on live conversations.
Conversations matched the customer, not the script.
- Sellence read tone, hesitation, and intent across the thread.
- Customers who'd hang up on a call replied to a question in their language.
Friction signals were routed, not buried.
- Regulatory question, policy team learned.
- Stalled objection, product saw the pattern.
- The thread closed deals and mapped where the business was losing them.

Why this transfers

Israel is the early signal. The pattern is global.

Libra's gain wasn't a quirk of a single market. The shift away from phone-led B2C sales is happening everywhere mobile reached saturation. The agencies adapting first are capturing the gap left by the ones still calling.

3-10%

US cold-call connect rate range, 2025. Down from 30%+ a decade ago.

39%

Of inbound calls to US insurance carriers go unanswered.

90%

Year-one insurance-agent attrition (LIMRA). Every retained agent compounds.

21×

Lift in lead-qualification odds when contacted within 5 minutes vs 30 (MIT Sloan / InsideSales).

High-consideration B2C, where the same playbook lands.

Auto & home insurance Health & supplemental coverage Life insurance Mortgage & refinance Home services & alarm High-ticket travel Wealth management onboarding Premium SaaS trial-to-paid

The next conversation

If your sales floor has the same problem, the playbook is portable.

30 minutes. We map your funnel, identify the conversation that is most likely to double, and tell you whether Sellence is the answer or not.

Book a 30-minute call → See more proof