There has been tremendous growth in the use of emerging technologies in the finance sector. Fintech startups have been upending the incumbents with innovative solutions. This week, we spoke to Pranshu Diwan, co-founder of GoPlannr, that provides ML-based solutions to insurance companies. This Bangalore-based startup wants to help insurance companies bridge digital design and efficient operations through the help of learnings in ML algorithms, behavioural sciences and automation
How Did It Begin?
GoPlannr was launched in October 2018. The idea hit the founders two years before they started GoPlannr, when we realized that the insurance industry is actually complicated and runs on outdated technology. Diwan’s journey began after he completed his MBA from IIM Calcutta, post which he joined a leadership program at Aditya Birla that was deciding to foray into health insurance. During this time, he realised that the insurance industry is not only complicated but also needs an efficient framework.
The fact that people had not made the best out of technology is what made them use it to augment a human adviser who will have a personal connect and reach through technology. According to the founder, the use of technology was a necessary step in order to simplify the three core problems of insurance which are complicated products, lack of trust and distribution.
Diwan said, “Building a good network, tech and data driven pipes is really crucial in order to capitalise the opportunities when the macro trends arise. Being able to simplify it as well as being able to cut down costs are the two key potential trends that the industry is going to follow and we are hopeful that we will be able to provide support when it actually happens.”
The Technology Behind GoPlannr
From a tech stack perspective, the product of the startup is an Android app which helps customers to book an adviser and a manager at a specified time. For instance, if an insurance company wishes to manage a field force of 80,000 agents, it is an understood fact that heavy expenditure will be occured in terms of manpower. In order to cut down the expenses, this app is created which will not only cut down cost but will also double their productivity. Since most of the work is actually done by the tech assistant, this will enable a single manager to handle 50-60 guys in one go.
There are about 4-5 areas in which machine learning has been used in their product. There are the following three core uses:
1.Sales teams can plan the perfect pitch with ML: The first application is inculcating ML in their pitches, since it is extremely important to propose the right plan in the right way. This product will ask the customer a series of questions like family details, members, income, occupation, health insurance. This is done to serve two purposes which is to figure out which combination of plans make sense to the customer and to make them understand the need of a protection gear. To serve this purpose, a bot has been created naming PRIA, an abbreviation for Product Recommending Insurance Assistant, which asks customers questions and also recommends a combination of plans. Combination of plans and customization are very important due to the presence of various segments of customers. It could be termed as a self-learning algorithm and not a static one. Here, machine learning is used to match the feature with the customer segment. Machine learning is however not applied in sales.
2.ML can enhance sales productivity: Machine learning algorithms are used to enhance sales productivity at a very personalized level. The aim is to organize a decentralized sales force in order to reduce dependency on management which directs people and creates a need for rigorous follow-ups. Typically, because of a management protocol people are not able to close sales in a certain segment. Machine learning not only provides solution for this by segmenting agents but also increases productivity. The cause of the risk is materialized due to machine learning which results in effective risk management. ML can assist in underwriting by providing price suggestions for different customers based on their individual risk factors.
3.Gain granular insights: Machine Learning here is basically helping the insurance company by providing them with the information about the potential customers and other insights that they don’t have access to, advisory services, data prediction and similar processes.
Challenges Faced
Initially, the biggest challenge the startup faced was in terms of adoption. They started off as a tech community. Their product was pitched to multiple insurance companies and also to independent agents. They later realised that drawing a map for the product and actually putting it to use are two very different things. Diwan said, “We are industry specific in terms of focusing only on insurance and we are leveraging technology to build better products but the issue is adoption. Therefore, to curb this problem, we’ve been using technology and insights to strengthen our key propositions and sales since the last 6 months.”
The startup believes that they are at a place where they can look at their competitors in two directions. The focus is to create a driving force to make community agents better in terms of competing techniques and distribution services. The first competitor technically will relate to the physical process aspect of it, which involves entering the data, calculating the applications and finally digitizing all of this in the app. GoPlannr is reinventing how a digital machine/equipment can make this process easier and better by smarter use of data, making clean and fast information available instantly on the phone.
Future Ahead
Going forward, the startup will work on solutions to simplify insurance products, help people buy the right plans quickly and ensure stress-free digital claims by digitising the insurance agent. They have built a smart robo-advisory app to guide and help improve the sales through training, recommendations, comparisons, FAQs, marketing tools and CRM on a tap.
GoPlannr plans to reinvent the entire insurance value chain using AI, behavioural science and tech with the following:
1) Product creation – using buying behaviours and preferences to co-create new products for different segments of customers
2) Underwriting – Auto-underwriting & dynamic pricing based on predicted risk
3) Claims – Fraud detection & Digital Claims adjudication
4) Distribution – opening up new channels (like WhatsApp, partnerships, etc.) for lead generation & closure using NLP
5) Operations – Improved customer experience at lower costs by replacing dependence on manpower for routine tasks
They aim to collaborate with 4-5 insurance companies within the next 6 months towards this goal. They are also aiming to open up their platform to independent agents to manage their own business across insurance companies.
