"Whether the industry is ready, consumers want to speak to their insurers through conversational AI, when and where they want to."
- Alain Desmier - Subcontext
Ask any insurance CEO to describe their AI strategy, and the talking points come quickly. Faster claims, document summarisation and a copilot for the call centre. Until earlier this month, the ambition for AI in protection from the C-suite was faster processing power; uncontroversial, understandable and safe.
Aviva, however, has fired the starting gun in a race that will come to define insurance distribution for the next decade. The significance of Aviva’s AI protection launch is not that consumers will suddenly rush to buy life insurance inside the ChatGPT app store, but rather that a major insurer has publicly acknowledged that AI is a planned route to market.
If an insurer can sell protection directly to customers via native AI, the race to create products that meet the demands of AI tools and AI-enabled customers is about to start. What does this mean for consumers, the quality of the protection product and the cost of distribution?
Close the conversation gap
Protection has a conversation problem. How do we protect people who never start the protection conversation or who do, but never come back, because filling in a simple quote form online creates a barrage of unwanted phone calls?
The protection industry has spent the last 20 years talking to itself, vehemently willing itself to be better. It spent years arguing about the telephone-based, advised vs non-advised sales process without noticing that an emerging consumer cohort, now in their 20s/30s, didn’t want either. Whilst consumers were busy transferring their pensions, opening bank accounts and buying investments, protection still needed '45 minutes on the phone.'
The result has been stagnation: annual policy sales are stuck at around 2 million. Younger consumers are now less likely to buy a life insurance product than ever before, and the FCA believes that 18 million consumers haven’t even considered protection products in any form.
AI will become a protection distribution channel
Online conversational interfaces are becoming the new front door to the financial services economy precisely because the consumer controls the timing, pace and channel of the conversation.
An insurer's native AI agent that can understand customer intention, retain the context thread of the enquiry and offer the consumer the chance to continue the chat on WhatsApp, SMS, or email, will at once transform the protection purchasing experience. Whether they are planning for it yet, creating the AI-native agent sales process will find its way onto every protection insurance roadmap within the next 12 months.
The strategic questions now become more interesting. Will insurers use AI to go direct under their own brands? Will they work with brokers to create single-tied products that can be launched quickly for specific distributors? Will they use AI to strengthen bancassurance partnerships by embedding effective conversational protection journeys where consumers already are? Or will they combine protection with savings, health and other financial products to create more rounded consumer propositions?
Shorter products will lower the cost of experimentation
One of the biggest barriers to growing protection is the cost of distribution and the manufacturing price of trying something new. It doesn't have to be this way. On stage at the Protection Review a few years ago, Jason Hurley noted that the UK has some of the world’s cheapest risk premiums, extremely low insurer margins, but the highest comparable broker commissions.
The cost of manufacturing and underwriting traditional 25-year term life insurance products, combined with legacy distribution technology and increasingly expensive lead generation, has created significant barriers to innovation. Privately, many insurers acknowledge that product reform is needed.
The average bought term life policy, underwritten for 25 years, is only held for 7 to 9 years, but it can take 2 to 3 years for an insurer to recover the indemnity commission paid to a broker at the point of sale. The result is a market built around long-term, inflexible products and distribution economics that discourage change.
Consumers increasingly want protection that is simpler, more flexible and able to evolve with their needs. If we can fix the 'front door' and make it easier for any broker or insurer to launch D2C, we can offer products that meet consumer needs based on what they, with unstructured conversational input, are telling us.
Winning will take bravery and disruption, in equal measure
Whether the industry is ready, consumers want to speak to their insurers through conversational AI, when and where they want to. Understanding their intention in language that they are comfortable with and that does not seek to belittle their lack of familiarity with protection language will unlock some of the 31 million consumers in the UK without a pure protection product.
'You’ll never change things, protection is sold and not bought' is the phrase any insurance innovator is used to hearing, normally at the conclusion of demonstration. Up to this moment, that was an unwanted side effect of the protection manufacturing process.
Now, with the power of AI tooling, we have a chance to reverse engineer the protection process and start with what the consumer actually wants, and price them from there.
