Increasing Retention in Insurtechs Through Churn Prediction
(2021) INTM01 20211Innovation Engineering
- Abstract
- Over the last decades, the Swedish insurance industry has seen decreased entry barriers due to deregulation and emerging new technologies, which have the potential to disturb the stagnated and consolidated competitive landscape of the industry. Initiated by newcomers like American insurance startup Lemonade, and later Swedish Hedvig among others, there is an increased push toward digitalization, transparency, and automation in the industry. This thesis examines how Insurtechs can increase retention by identifying customers at-risk of churning, as well as what actions they can take in order to make customers more likely to stay, with the digital insurance company Hedvig as a case study. Various machine learning methods for predicting churn... (More)
- Over the last decades, the Swedish insurance industry has seen decreased entry barriers due to deregulation and emerging new technologies, which have the potential to disturb the stagnated and consolidated competitive landscape of the industry. Initiated by newcomers like American insurance startup Lemonade, and later Swedish Hedvig among others, there is an increased push toward digitalization, transparency, and automation in the industry. This thesis examines how Insurtechs can increase retention by identifying customers at-risk of churning, as well as what actions they can take in order to make customers more likely to stay, with the digital insurance company Hedvig as a case study. Various machine learning methods for predicting churn are examined in a literature review, and a model is developed and proposed for Hedvig. Seven levers for increasing retention, 1) Understanding Churn, 2) Customer Intake, 3) Product Improvement, 4) Lock-in, 5) Targeting at-risk Churners, 6) Save Desk, and 7) Organizational Setup, are identified and presented with documented best practices from expert interviews. The conclusion is that churn could not be predicted accurately as the proposed model, a Gradient Boosted Tree model, achieved an ROC value of 62%, which is considered low, and an unsatisfactory precision and recall curve. In the discussion section, we propose that the reason behind this is that there is that there is not enough signal in the data, that the two classes are very homogeneous. In order to improve the predictive accuracy, more usage data from the customers, that have a stronger correlation with the outcome variable, churn, should be collected. Besides predicting churn, the thesis discusses some alternative ways to increase retention, based on discussions with industry professionals, and presents some company specific recommendations in the discussion chapter. (Less)
- Popular Abstract
- Over the last decades, the insurance industry has seen decreased entry barriers due to deregulation and emerging new technologies, which is expected to disturb the stagnated and consolidated competitive landscape of the industry. Insurtechs - insurance companies leveraging new technologies - push the digitalization, transparency, and automation in the insurance industry, which is ripe for disruption.
Insurance as a product has existed for a long time, with traces dating back to risk-sharing contracts in Babylon 300 BC. Traditionally, the Swedish insurance industry has been strictly regulated, which has led to a high level of consolidation. However, the industry has recently seen an increased degree of innovation, partly driven by new... (More) - Over the last decades, the insurance industry has seen decreased entry barriers due to deregulation and emerging new technologies, which is expected to disturb the stagnated and consolidated competitive landscape of the industry. Insurtechs - insurance companies leveraging new technologies - push the digitalization, transparency, and automation in the insurance industry, which is ripe for disruption.
Insurance as a product has existed for a long time, with traces dating back to risk-sharing contracts in Babylon 300 BC. Traditionally, the Swedish insurance industry has been strictly regulated, which has led to a high level of consolidation. However, the industry has recently seen an increased degree of innovation, partly driven by new entrants, and it has been argued that insurance is experiencing a similar ‘wave of disruption’ as that of the post-2008 financial industry. Initiated by newcomers like American insurance startup Lemonade, and in Sweden by Hedvig, there is an increased push toward digitalizing insurance and to improve efficiency and automation and increase customer satisfaction in the industry.
The environment for new innovative companies to succeed in the insurance industry is optimistic: advancements in digitalization, information technology, and machine learning could be highly valuable for Insurtechs - insurance companies taking advantage of the tools provided by the digital transformation, including predictive modeling.
As insurance is based on a subscription business model and is in essence reliant on recurring revenue from premiums, it is important for insurance companies that a large share of their customers keep their contract for a long period. Adding new customers to the platform comes with an acquisition cost in the form of marketing and the operational costs of onboarding new users. This makes it important to actively work to retain existing customers. Studies have shown that the cost of acquiring a new customer is about ten times higher than to retain an existing customer, and an increase in retention from 85% to 90% can give rise to an increase in net present value profits from 35% to 95%.
The method of increasing retention in focus in our study is to target customers who are at-risk, or likely to cancel their contracts, with directed actions. Focusing on the customers who are actually at risk, rather than targeting all customers, has many advantages including not wasting the marketing budget and ensures that one ‘lets sleeping dogs lie’, not interfering with customers that have ‘forgotten’ about their subscription. In order to target these customers effectively, one first needs to identify which customers are at the doorstep of leaving. There are varying ways of identifying these customers with different levels of sophistication: from simple rules based on experience to more recent predictive methods based on advancements in machine learning.
Our study shows that the importance of retention management is increasing, as it is becoming more expensive to acquire new customers and they have more complex needs. In 2021, the spend on existing customers is expected to increase by 30%, partially due to data availability and new technology. These investments are often spent on new digital technology such as marketing automation, in order to better handle the customer relationships. Proactive retention work, combined with reactive save desk operations, can have a large effect on retention, and can decrease the churn rate by between 25-50% in Insurtechs.
In order to make accurate predictions, it is critical to collect and develop trigger’ features in the data set that strongly signals that a user is on the verge of terminating their insurance, as even the most advanced machine learning models do not add any value if the right data is not being collected. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/9055369
- author
- Rapp Farnes, John LU and Christiansen, Oskar LU
- supervisor
- organization
- course
- INTM01 20211
- year
- 2021
- type
- H2 - Master's Degree (Two Years)
- subject
- keywords
- Non-life insurance, Property and casualty insurance, Customer retention, Churn prediction, Predictive analytics, Classification, Machine learning
- language
- English
- id
- 9055369
- date added to LUP
- 2021-06-17 15:17:14
- date last changed
- 2021-06-30 08:42:34
@misc{9055369,
abstract = {{Over the last decades, the Swedish insurance industry has seen decreased entry barriers due to deregulation and emerging new technologies, which have the potential to disturb the stagnated and consolidated competitive landscape of the industry. Initiated by newcomers like American insurance startup Lemonade, and later Swedish Hedvig among others, there is an increased push toward digitalization, transparency, and automation in the industry. This thesis examines how Insurtechs can increase retention by identifying customers at-risk of churning, as well as what actions they can take in order to make customers more likely to stay, with the digital insurance company Hedvig as a case study. Various machine learning methods for predicting churn are examined in a literature review, and a model is developed and proposed for Hedvig. Seven levers for increasing retention, 1) Understanding Churn, 2) Customer Intake, 3) Product Improvement, 4) Lock-in, 5) Targeting at-risk Churners, 6) Save Desk, and 7) Organizational Setup, are identified and presented with documented best practices from expert interviews. The conclusion is that churn could not be predicted accurately as the proposed model, a Gradient Boosted Tree model, achieved an ROC value of 62%, which is considered low, and an unsatisfactory precision and recall curve. In the discussion section, we propose that the reason behind this is that there is that there is not enough signal in the data, that the two classes are very homogeneous. In order to improve the predictive accuracy, more usage data from the customers, that have a stronger correlation with the outcome variable, churn, should be collected. Besides predicting churn, the thesis discusses some alternative ways to increase retention, based on discussions with industry professionals, and presents some company specific recommendations in the discussion chapter.}},
author = {{Rapp Farnes, John and Christiansen, Oskar}},
language = {{eng}},
note = {{Student Paper}},
title = {{Increasing Retention in Insurtechs Through Churn Prediction}},
year = {{2021}},
}