Looking back over the past 25 years, my favourite role hasn’t been the CEO, COO (Chief Operations Officer) or CTO (Chief Technology Officer) roles you’d expect me to say, but rather the humble Business Analyst role I first held at Goldman Sachs when that job title was still waiting to be adopted this side of the Atlantic. With a focus on documenting both our current and target business processes, the role offered that perfect combination of thinking strategically but creating value tactically, as well as working across the business and IT rather than working inside one of those silos. I could think of no other role that was as well placed to make a meaningful difference to how business was actually ‘done’.
It wasn’t long before I migrated from Visio (or today’s Miro) to using the industry standard for modelling processes (BPMN 2.0) where you can model the process in such a precise and nuanced way that you could import the model into a BPM (Business Process Management) system like Camunda to push the process ‘live’. This was the original no-code before no-code became a thing. It’s a shame it never went mainstream outside of Germany, with the notable exception in the UK of my former employer Goldman Sachs.
It’s also a technology that rather than become obsolete has continued to remain more relevant than ever, although as insurers and TPAs move from being product-centric to customer-centric, my sense is that we need to stop thinking in a ‘left-to-right’ way on how the process flows across you, your customer and your supply chain, and start thinking more creatively in a ‘right-to-left’ (i.e backwards) on what business decisions are preventing you from instantly resolving the customer’s need. This is the conceptual difference between those who are still thinking in terms of eNOL or eFNOL with its focus on notification (the first step in a linear product-centric process) rather than the paradigm shift of eROL (eResolution of Loss) with its focus on immediate zero-touch resolution (the last step in a customer-centric journey)
Practically, this means thinking more in terms of decisions rather than tasks. For example, in a digital-first claims process, high-level decisions might be; Is the person completing the claim who they say they are? Do we have all the data to process the claim? Is the uploaded document the document we asked for? Is there any suspicion of fraud? Are we ok to pay?
Once you have listed all of the business decisions you need to make to approve the claim, you can start to document the logic for each business decision, and identify whether there is a hierarchy of decisions. For example, “Are we ok to pay” is probably the highest level decision that is based on the answer to all of the other decisions mentioned above. For those looking to develop their technical skills, I would strongly recommend reading Bruce Silver’s Method & Style books and the use of Signavio to start modelling both decision tables and your thought process for decision-making.
Understanding which decisions your process is dependent on can also accelerate the appropriate choice of technology. Let’s say you want to take a digital-first approach to automating the “Is the person completing the claim who they say they are?” decision. One challenge insurers face is that unlike banking, the customer is unlikely to have your app on their phone so you can’t make use of TouchID to authenticate them, and asking the customer to download an app to make a claim is a no-go. Many insurers are now using Claim Technology’s no-code chatbot and comparing the date of birth, postcode etc entered by the customer to information already present on the policy. But to me this seems susceptible to fraudsters/identity theft, so why not explore how voice notes and Passive Voice Authentication technology, from a start-up like Blockfrauds might help authenticate customers silently in the background without the customer going through a formal authentication step. By understanding the individual decisions that need to be made, I feel that you then have much more freedom to think creatively about how to optimise the customer experience, rather than feel straight-jacketed by a sequential process.
The shift to thinking in terms of decisions also dovetails nicely into the work companies are already doing around the application of AI to automate decisions, leveraging insurtechs such as Cortical, Omnius and Sprout.ai, all of whom can be accessed via Claim Technology’s insurtech marketplace, as well as the governance and compliance work companies will need to do to meet the regulatory requirements of Europe’s much publicised AI Law. Interestingly, insurtechs offering certification around the use of AI can also be found in Claim Technology’s insurtech marketplace.
If we can take a decision-first rather than process-first approach to modelling business processes, I also think that it becomes an awful lot easier to then look at adopting BPMN to orchestrate those real-time decisions.
Developing world-class process and decision modelling skills is vital for industries wanting to excel in the AI economy. Isn’t it time we adopted these skills in the UK to attract and retain the next generation of talent? If you’re up for that, ping me at hello@claimtechnology.co.uk.
