Solution · Insurance · renewals

Customers who may not renew, found automatically weeks before the renewal date

We build this system to order for insurers. Every day ordinary software lists the policies coming up for renewal and sends reminders on your schedule, and AI reads each customer’s payments, claims and messages, weighs whether they may not renew and why, and picks one of the offers you have approved. Customers with a complaint, an open claim or a big account go to a retention manager

Try an exampleBook a call (30 min)

Who works with it

Retention manager

The insurer’s employee who works with customers who may leave

Gets the customers a person should talk to, each with what AI weighed and why, and approves offers outside the approved ones

Head of sales

Runs sales and renewals

Sets the reminder schedule, the offers AI may pick from and which customers always go to a person. Sees what was sent to whom and why

Policyholder

The customer whose policy comes up for renewal

Gets a reminder or an offer in the channel they answer, before the policy ends rather than after

What comes in

What the system reads every day, and what it sends

Policy systemdates, coverageCRMcustomer, contactsPaymentson time or lateClaims historyclaims, complaintsMessagesemails, chats, surveysThe daily reviewreads, changes nothingReminders, offers, tasksto the channels and the CRMAIAIPolicy systemdates, coverageCRMcustomer, contactsPaymentson time or lateClaims historyclaims, complaintsMessagesemails, chats, surveysThe daily reviewreads, changes nothingReminders, offers, tasksto the channels and the CRMAIAI
Three ways out

The path of one renewal

Policy due for renewalfrom the policy system
AIThe daily reviewrules and AI
  1. Reminders onlyno sign they may leave
  2. An approved offersent by the system
  3. Retention managercomplaints, big accountsa person decides
Example customers

Pick a customer and see the plan before the renewal date

Made-up customers. Their days, channels and offers stand for the rules an insurer would set

Pick an example customer

Home insurance, the premium went up this year

The plan before the renewal date
  1. 45 days beforeOffer: monthly installmentsEmail · the system
  2. 21 days beforeThe offer againMessenger · the system
  3. 7 days beforeLast reminderSMS · the system
  4. Renewal date
What AI weighed AI
  • May leaveAsked in the chat why the premium went up
  • May leaveAsked whether the premium can be paid in parts
  • KeepsNo claims in three years

AI’s reading: may leave over the price

What happens next AIAn approved offer

AI sees a reason the customer may leave and picks the approved offer that answers it. The system sends it on the insurer’s schedule

The offer: paying the premium in monthly installments

Software, AI and a person

Who does which part of the work

Software and AI go through every renewal. A person takes the customers who need one

Rule

Ordinary software

Monotonous work with clear rules

  • Lists every policy coming up for renewal, every day
  • Sends reminders on the schedule the insurer sets
  • Checks each offer against the insurer’s rules: who qualifies and within what limits
  • Sends open complaints and big accounts to a retention manager
  • Logs every message and answer in the CRM
AI

AI: understand, weigh and choose

The small decisions on every renewal

  • Reads payments, claims, complaints, chats and survey answers
  • Weighs whether the customer may not renew, and why: the price, a claim, or simply missing the date
  • Picks the approved offer that answers that reason
  • Writes the message and picks the channel the customer answers
  • Chooses whether a reminder is enough or a person should talk to the customer
a person decides

A person

Every conversation and every exception

  • The retention manager talks to customers with a complaint or an open claim
  • The retention manager handles big business accounts
  • A person approves each offer outside the approved ones
  • The head of sales sets the offers, the limits and the schedule
  • Only a person changes a price or coverage beyond the rules
How a project starts

Four stages. After each one, you decide whether to go on

First on last year’s renewals, then on new ones

  1. Usually firstDiscovery and specification
    In
    How renewals are handled today, the history you keep and the channels you use
    You get
    A written specification: what the system weighs first, what data it needs, what it will cost and in what order. The document is yours whether or not we build it
  2. ThenTest on past renewals
    In
    Last year’s renewals: who renewed, who left, and their history before the date
    You get
    The customers the system would have flagged, next to who actually left
  3. ThenPilot on one segment
    In
    Renewals in one line of insurance or one region you choose
    You get
    Reminders and offers for that segment. A manager checks each offer before it goes while the rules are being set
  4. When you decideRollout
    In
    The lines and regions you choose
    You get
    Every renewal there goes through the daily review. Your team changes the offers and the schedule
Six questions we ask first
  1. How many policies come up for renewal a month? We plan the system and the pilot around it
  2. What share of policies do not renew today, and do you know why? The test on past renewals starts from it
  3. Which channels do you use: email, SMS, messengers, calls? The system sends through the ones you already have
  4. Do you keep a history of contacts with each customer? AI reads it to see why a customer may leave
  5. Which offers can a retention manager make today? They become the approved offers the system picks from
  6. Which CRM does your sales team use? The pilot connects to it, and every message is logged there

Tell us how renewals are handled today

A 30-minute call. We will say what your customer history allows and where a project would start