Stop automating bad processes

I have increasingly found myself wondering whether we should still be talking about “technology” as a separate subject in insurance.

Technology has become so embedded in virtually everything an insurer, broker or service provider does that perhaps the more useful conversation is about business: how we operate, how we connect, how we serve customers and, ultimately, how effectively we execute.

Technology has become so embedded in virtually everything an insurer, broker or service provider does that perhaps the more useful conversation is about business: how we operate, how we connect, how we serve customers and, ultimately, how effectively we execute.

That is an important observation. Insurers and brokers today have access to APIs, artificial intelligence, sophisticated analytics, digital payment capabilities, cloud infrastructure and an expanding universe of specialist technology providers.

The problem is no longer finding technology. The problem is making everything work together. Many insurers continue to operate on technology environments and operating models built up over decades. New systems have been added, integrations created, processes adapted and layers of complexity accumulated. The result is often an impressive collection of technologies that do not necessarily create an impressive customer experience.

Dave described it as “an organisational execution challenge”, rather than a technology challenge. I think that distinction matters enormously. Too often digital transformation focuses on improving one small component of a process. We automate a task, digitise a form or connect another system. Each improvement may be useful, but customers don’t experience insurance as individual processes.

They experience a journey. The real opportunity therefore lies in redesigning that journey end-to-end, with the policyholder at the centre. Premium collections provide an excellent example. Collections infrastructure is something most customers probably never think about when it works properly. But the moment something goes wrong, it becomes extremely important.

As Dave said: “It sits in the background until it doesn’t work. And then you’re talking about money.” A failed premium collection, delayed refund or incorrect payment can very quickly transform an invisible administrative process into a customer experience problem. Modern collections therefore need to mean much more than successfully processing a debit order.

Customers increasingly expect multiple payment options, real-time notifications and visibility over what is happening. Brokers want less administration and a single, clear view of a client’s payment status. Insurers want improved collection success and reduced reconciliation effort. Achieving all three requires the ecosystem to work together.

That word, ecosystem, kept resurfacing during our conversation. It is also where I believe one of the biggest opportunities for insurance lies. Every participant in the insurance value chain possesses part of the information required to understand the customer. The insurer has one set of data, the broker another, while payment and collection providers have transactional information.

Individually these are pieces of data. Connected, they become intelligence. Consider something as simple as a premium repeatedly failing on the 25th of every month and succeeding when collection is attempted again several days later. The technology can identify the pattern. But identifying it achieves very little if nobody changes the collection date.

That perfectly illustrates the difference between having data and actually using it. As Dave put it, the challenge is: “How do we turn that data into information and knowledge?”

Artificial intelligence will undoubtedly accelerate this ability. Fulcrum itself is taking a cautious approach, using AI practically to support development, exception handling, reconciliation, response times and operational efficiency. Interestingly, the early benefits are not necessarily spectacular AI applications. They are improvements in speed, quality and time to market.

Perhaps that is exactly how we should be thinking about AI. Not as something that needs to produce a dramatic headline, but as another increasingly powerful tool for making businesses work better.

Looking towards the next five years, Dave believes insurers and brokers should invest in API-first architectures, strengthen their data foundations and simplify processes. But there was one piece of advice from our conversation that stood above everything else for me: “Don’t just automate a bad process. Fix the process.”

It sounds obvious. Yet organisations frequently take inefficient processes, automate them and then celebrate the resulting efficiency. All they have really achieved is executing a bad process faster.

Dave recalled being asked to automate one such process and responding: “Why? So that we can just break it faster?”

Therein lies perhaps one of the most important lessons for insurance transformation. Before buying another system, deploying another AI tool or automating another workflow, we should first ask whether the process itself makes sense. Remove unnecessary steps. Connect the right information. Decide what outcome we are trying to achieve. Understand the customer journey. Then determine where technology can improve it.

Because the next competitive advantage in insurance will not come from simply having more technology. Almost everyone will have access to the same technology. The advantage will come from connecting technology, data, people and processes across the insurance ecosystem in a way that removes friction for the person who ultimately matters: the customer.

Technology will continue to become faster, smarter and more capable. The differentiator will be how well we use it.

Dave Stevens
Chief Technology Officer
Fulcrum Collections