Revolutionising Insurance Marketing: Automation, Personalisation, and Data-Driven Strategies at PolicyBoss
In this interview, Varun Kaushik, Executive Vice President and Head of Marketing at PolicyBoss.com (Landmark Insurance Brokers), shares his journey of transforming PolicyBoss from a traditional insurance broker to a B2B2C model driven by marketing automation, personalisation, and advanced technology. He delves into how the company has adapted to regulatory changes, embraced inbound marketing, and introduced innovative tools like AI-powered onboarding and “Live Brokerage” to boost productivity and retention. Varun also discusses his strategic approach to campaign management and automation, offering valuable insights and best practices for creating impactful, data-driven marketing strategies in highly regulated industries like insurance.
Key Takeaways
- PolicyBoss has undergone significant digital transformation, pivoting from a B2C to a B2B2C model with 85%+ online business.
- AI-driven conversational bots and personalised features like live localisation and instant commission payments have boosted daily policy sales from 500 to over 1500.
- Mobile push notifications and WhatsApp have become primary channels for customer engagement, replacing email and SMS for promotional content.
- Personalisation, intent mapping, and unified customer journeys are crucial for effective campaign management and automation.
“A common mistake is creating separate journeys for marketing, retargeting, and re-engagement. In today’s converged market, these need to be part of one cohesive, unified consumer journey with multiple stages.”
Please share a brief overview of your professional journey, which led you to your current position as the Executive Vice President and Head of Marketing at PolicyBoss. What are your primary responsibilities in your current role, and how have these responsibilities evolved since you joined the organisation?
Varun: We’ve pivoted our entire business model at PolicyBoss, which was originally a brick-and-mortar insurance broker. Up until 2017-2018, our focus was primarily B2C. However, with the new POSP (Point of Sale Person) guidelines introduced by IRDAI (Insurance Regulatory and Development Authority of India), we saw an opportunity and shifted to a B2B2C model. Today, over 80-90% of our business is agent-driven, with 85%+ of that conducted online through our own digital platforms. This transformation has been underpinned by significant advancements in our back-end systems, including LMS (Learning Management System), CRM (Customer Relationship Management), automation, and even RPA (Robotic Process Automation) for reporting and analytics.
Our in-house platform, which started as a simple tool to meet internal needs, has evolved into the core middleware for our entire agency business. Today, it not only supports our website and app but also serves as our LMS, CRM, ERP, and back-office management platform for online business.
Our journey now continues with the integration of generative AI. We’re currently building out an AI-powered, fully digital onboarding process that requires zero manual intervention. This is a complex project, involving three external partners and over 200 possible conversation flows. The 11-step onboarding journey aims to create a natural, human-like experience.
On the automation front, we’re implementing features to enhance both agent acquisition and productivity. For example, our industry-first “Live Brokerage” feature displays potential commission earnings in real time for agents, while our instant commission payment feature “InstaPay” eliminates need for reconciliation completely and also narrows down the typical 45-day wait, allowing agents to receive their commissions immediately after issuing a policy. These innovations have had a huge impact: since launching these features in June 2024, our daily online policy volume has grown from 500 to over 1,500.
In summary, this transformation not only reshapes our business but also sets new standards in distribution, creating a seamless and confident experience for our producer community. These advancements are as relevant to any distribution-based business as they are to insurance broking, underscoring the universal need to enhance communication and support for distribution networks.
How do you develop and implement marketing strategies at PolicyBoss? Can you provide insights into your approach to managing multi-channel campaigns, including digital and traditional media?
Varun: COVID-19 significantly reshaped consumption patterns in the insurance sector. For a couple of years post-pandemic, health insurance saw a surge in interest as people recognised it as a critical safety net against unforeseen risks. Life insurance also experienced a notable increase, driven by a sense of urgency and FOMO.
Interestingly, vehicle insurance did not experience as dramatic a shift in demand. However, health insurance has grown so substantially that it now rivals—if not surpasses—vehicle insurance, which has traditionally been the largest subcategory in general insurance.
At PolicyBoss, we have closely tracked these shifts in consumer behaviour. Today, the approach can no longer be “one-size-fits-all”; instead, it demands nuanced strategies tailored to specific products, sub-products, regions, and even income levels. By leveraging behavioural data, we can now model consumer propensity with 60-80% accuracy, helping us predict needs and optimise offerings.
While there is greater awareness and an increase in household spending on health and life insurance, overall, India remains underinsured. This calls for more than just growth in the industry—it requires deeper, proactive engagement and strategy adjustments to adapt to rapidly changing consumer behaviours.
Campaign planning now involves constant analysis of behavioural trends, which are evolving every month. Flexibility is essential, with Plan B always ready in case the primary strategy falls short. With the increased availability of actionable data, campaign management has become more precise and data-driven. However, driving widespread adoption remains a challenge. Unlike in other categories, where consumers quickly embrace discounts and incentives, insurance still needs greater momentum to achieve the same traction.
Read More
From a marketing automation perspective, how do you incorporate industry-specific norms into your strategy? Given that PolicyBoss is an insurance broker in a highly regulated sector, how does your approach differ from industries like retail that face fewer restrictions?
Varun: Being an insurance broker, our responsibilities and guidelines differ significantly from those of insurance companies, as we’re governed by distinct regulations. Over the past seven years, these marketing guidelines have undergone frequent revisions, balancing leniency in some areas with stricter controls in others. Recently, the telecom authority’s efforts to curb spam have prompted us to shift to a low-spam model, a transition we embraced early on.
At PolicyBoss, we’ve made a significant pivot from an outbound-driven to an inbound-driven model. This shift requires adapting our consumer journeys accordingly. For instance, our website includes robust data privacy policies, emphasising our commitment to handling sensitive personal information with the utmost care—a critical priority in our industry, where misuse can have significant repercussions.
In terms of automation, once we’ve crafted the right messaging and tone, automation becomes the tool for execution. The foundational events in the insurance buying journey—search, quotation, decision-making, and payment—haven’t changed. However, the tone and clarity of our communication at each stage have evolved. Insurance marketing as a whole has become less cluttered, with a focus on simplifying information for consumers.
A common barrier for consumers is the belief that insurance involves too much fine print. Surprisingly, this is often the same for other financial products like futures, options, or equities, which carry greater risk. Yet insurance uptake remains lower due to perceived complexity and unclear language. Addressing this, we’re leveraging automation and adaptive, intelligent advisory tools to provide real-time tooltips, enhancing both agent and customer understanding.
For example, car insurance add-ons have expanded in recent years. Without clear explanations, customers may drop off when encountering unfamiliar options. We’re tackling this by embedding user-friendly guidance to make the process simpler and more intuitive. Automation orchestrates the communication flow—such as video journeys or successive reminders based on user actions or inactions—while these nuanced interventions build confidence and clarity in the process.
All these improvements in experience are targeted at the POSP agents that partner with us as they are the first point of contact for policyholders. Educating them, empowering them and building confidence within them has a trickle-down impact on end customers as the more in command these PoSPs would be, the more their end customers are bound to be sure and swift in their insurance-buying specific considerations and decision-making.
Ultimately, while the insurance purchasing process is uniform across brokers, companies, and web aggregators, few provide the clarity consumers need without prompting. Solving this is our focus, leveraging automation to create a streamlined, supportive, and insightful experience for every user.
How are you currently mapping intent within your data set after running a campaign, and how do you segment your audience based on their intent?
Varun: Our approach to intent mapping is a three-step, interconnected process built into our homegrown system, which assigns a numerical intent score.
- On-Page Behaviour: Both agents and customers interact with our website and mobile app across different product journeys. Intent scoring begins with analysing the user’s stage in the journey and their activity via heat maps, which highlight the areas of the page where they spend the most time. This combination of page interaction and time spent helps define their on-page intent score.
- Purchase Intent: Purchase intent is gauged starting from the insurance provider selection stage. Users who reach the proposal stage or beyond exhibit a higher intent than those still in the search or comparison stages.
- Historical Data & Baselines: With 20 years in the business, we leverage historical data to establish baselines for each period (e.g., daily, monthly, or fortnightly). This allows us to compare current performance against established averages to refine scoring.
Based on this multi-factor scoring system, we’ve set up automated communication journeys for high and low intent. For higher-intent users, the focus is on nudging them toward conversion with reminders and prompts. For lower-intent users with high potential, the messaging is more strategic and layered, gradually encouraging them toward closure.
With multiple channels available for lead acquisition, both inbound and outbound, where does email stand in terms of effectiveness for you? What would you say are your top three to five most effective channels today?
Varun: On the retail side, customer preferences are increasingly leaning towards completing their transactions seamlessly on their mobile devices, without navigating through multiple screens. Convenience is now a critical factor across retail, not just in insurance. This shift is evident in the rise of quick-commerce platforms like Swiggy and Zepto, highlighting how mobile is no longer a secondary screen—it’s the primary one.
At PolicyBoss, we’ve aligned our strategy to this trend. For logged-in users, push notifications have become one of our most effective communication channels. Initially, I was sceptical about the impact of mobile notifications on purchase decisions, but after observing how new age brands like Zomato, Swiggy and Zepto use them, we adopted a similar approach. By identifying peak and non-peak usage times, we now send targeted push notifications during non-peak hours, using action or income prompts to engage our agents effectively.
For promotional messages, WhatsApp has become our top-performing channel. Email and SMS, once key channels, are now primarily reserved for transactional notifications due to their high clutter. With inboxes overwhelmed by promotional emails from various brands, being a needle in a haystack doesn’t serve our needs in a low-conversion category like insurance.
Human behaviour supports this shift: while many tolerate hundreds of unread emails, an unchecked WhatsApp notification is much harder to ignore. Though more expensive than SMS, WhatsApp’s higher engagement rate justifies its use. However, we selectively qualify users for WhatsApp to avoid spamming, ensuring our brand numbers aren’t marked as spam. This approach has yielded substantial results, especially in targeted campaigns.
For example, we recently ran a WhatsApp renewal campaign for luxury car owners, personalised with their name, car model, insurance expiry date, and renewal options. Tied to the theme of a “worry-free Diwali,” we encouraged them to renew before the holiday. The campaign led to a 72% renewal rate among customers, with policies expiring on Diwali, a result we hadn’t seen in previous efforts.
WhatsApp’s effectiveness lies in its relevance and meaningful messaging. Messages must be action-oriented and specific. A successful WhatsApp campaign is not a mass broadcast—it’s a direct, purposeful engagement. If a message is too preachy or generic, it’s likely to get blocked. In this channel, relevance is paramount, and action-driven content is essential to maintain engagement.
How do you use personalisation in your marketing campaigns to enhance customer trust and engagement? Could you share some best practices for defining an effective personalisation strategy?
Varun: Having a clear retention strategy is essential, especially when planning around established benchmarks. A skilled marketer understands the key levers of conversion, and open-ended messaging won’t cut it. The messaging must be highly relevant and concise.
For instance, in hyper-local markets, using the appropriate local language is crucial. Sending a Hindi message to a region where it’s not commonly spoken is ineffective. Similarly, using overly complex language or superlative terms just to sound impressive often backfires, leading to customer disengagement. Brands, particularly in the service industry, should adopt a tone of humility in their communications.
Once the messaging is refined, it’s important to anticipate potential points of failure and build a journey that accounts for these, with retargeting and re-engagement as integral parts of the overall strategy—not isolated components. A common mistake is creating separate journeys for marketing, retargeting, and re-engagement. In today’s converged market, these need to be part of one cohesive, unified consumer journey with multiple stages.
Without this unified approach, campaigns are fragmented, and it becomes challenging to gain a holistic view of the customer journey. By integrating these stages into a single, seamless journey, marketers can more effectively monitor, analyse, and optimise for retention, ensuring all potential failure points are addressed and gaps are minimised.
You run various types of campaigns—awareness, promotional, and transactional, each with different goals like conversions. How do you measure the ROI of these campaigns, and what role does lead attribution play in maximising ROI and ensuring accurate measurement?
Varun: For us, ROI or ROAS is clearly measured by policy sales, including the number of policies sold and the premium value compared to targets. We also track renewal conversions and premium values, making it straightforward to quantify performance since it’s all conversion-based.
The challenge, however, lies in attributing the reasons for missed conversions. In India, there’s still a cultural reluctance to openly share why a potential customer didn’t proceed with a purchase. Despite running surveys, response rates remain low—around 10%.
While conversion attribution provides a clear picture of our business performance, understanding the ‘why not’ behind missed opportunities remains an ongoing challenge. This is how we assess and evaluate the success and areas for improvement in our campaigns.
For the campaign management and automation report we’re creating, what insights do you believe would add the most value for readers? Also, could you share some best practices from your own experience that you think would be valuable for others reading the report?
Varun: I believe one of the most critical factors in marketing automation is knowing precisely what to automate and how to market it effectively. Success doesn’t come simply from choosing a top-rated service provider; it requires clear objectives and a deep understanding of your own business needs. If you expect a service provider to handle both discovery and implementation without your clear guidance, that’s a shallow approach.
You need to make the necessary investments to be well-prepared, even if that means upskilling or bringing in expertise. Not everyone has in-depth knowledge of marketing automation’s technical side, and there’s no shame in that. With most providers now operating in a cloud-based environment, REST APIs and/or webhooks are standard for integrations, and many tools can handle revenue modelling directly within the system, eliminating the need for separate spreadsheets or accounting software. Another notable change is today’s integrations factor a higher concentration of push calls from marketing automation systems and more complex journeys see the deployment of PUSH as well as GET calls, especially in case of 2-way automated communications.
Before you begin the automation journey, establish concrete, measurable objectives. It’s essential to have tangible metrics—not just projections or hypotheses. Use either your own performance data or industry benchmarks as a baseline. Without these numerical milestones, marketing automation will likely fall short, and as a marketing leader, you’ll struggle to demonstrate its value.
A second key consideration is maintaining flexibility and clarity in event definitions and consumer journey mapping. Many marketing automation conversations stall due to a lack of clarity in these areas. Be open to experimentation—even if you fail, the learning process is invaluable. In today’s fast-paced world, it’s sometimes necessary to test multiple strategies simultaneously across the same audience segment. Choose a service provider who understands your industry and business well; if both parties are learning through discovery, measuring impact becomes difficult, and results will likely feel suboptimal. A best practice I strongly recommend is building cohorts to simplify reading and acting upon sub-segmented behavioural trends. Another suggestion is to always take a funnel view for all campaigns, as “for information only” does not apply to today’s marketing environments. Every marketing rupee spent, needs to be attributed towards the desired impact and should have a number value, not a notional one.
Finally, the tone of voice and metrics used in campaign management and automation should be tailored to each campaign and line of business. From my experience in both B2B and B2C, I’ve found that approaches like account-based marketing (ABM) can be adapted for both. In B2B, ABM targets specific accounts with predefined values, activating an entire marketing engine around these key accounts. This principle can be applied to B2C as well, especially when targeting high-value transactions or luxury items. Flexibility in campaign design, tailored to the order value and customer profile, is crucial for success.
Lastly, I strongly recommend that for any and every comm outreach that the brand makes, a control group should be created to exclude a 2-5% audience and see a degree of variance in behavioural / action-oriented response between those who received comm and those who didn’t.
These are my key insights on the subject.
