Advice for insurers: put the decision before the data

July 2016 | EXPERT BRIEFING | RISK MANAGEMENT

financierworldwide.com

We live in the age of Big Data. For insurers, the extent of data access now available through public reports, Google Earth or data mining is astonishingly helpful. If you are writing a policy for a property in a flood plain area, for instance, you have the meteorological reports for the last 15 years at your fingertips to inform your decision. Thanks to the latest technologies, the current (and growing) level of data availability means you can see what’s happening in real-time better than ever before, and make decisions based on this. Better still, such extra information can be tailored for each claim to evaluate risk, inform how to price insurance products, or allow eligibility in traditionally challenging areas.

However, data is a means to an end, not the end itself. Too many businesses have learned the hard way that a Big Data project team shouldn’t start by collecting data – it should start by focusing on the decision or decisions that need to be made, and only then consider the most appropriate data to be used to better inform that decision. Furthermore, insurers need to have ‘decision discipline’ which ensures that they not only use data to make better decisions, but that they record, analyse and modify those decisions to continually improve performance.

The job of the actuary continuously changes and is challenged daily by the increase in data availability, so insurers must develop the agility to deal with rapid changes in policyholder behaviour, regulations, market conditions and new technologies. This involves not just using Big Data, but recording the fruits of data-based decisions in a transparent and auditable form, because real business value lies in the decisions that an organisation makes on a day-to-day basis. It’s not enough to make great decisions, though – insurers need to treat these decisions like the valuable assets they are and store them for future learning to facilitate continuous improvement.

Historically, like many industries, the insurance industry has tended to be quite poor at centralising, managing and recording decision logic. This is not through lack of trying or operational expense incurred. For instance, each time a team has to decide how much to underwrite an oil rig for, most would start from scratch or have to flip through files from the last similar case. By embracing technology, insurers can save the back-breaking paperwork and glean useful decisioning information to get a better handle on the risks by accessing company data online via a mobile device. Recording intellectual property in this way, and understanding what’s working, how and why, is invaluable for future commercial success – especially in an industry where the reasoning behind decisions is so crucial to the product on sale.

In many industries, regulators expect companies to be able to explain how and why certain decisions were made, so having a structured system of all claims decisions easily to hand is almost reason enough to move with the times when it comes to decision management. If you design a model, test it, execute it and record the outcome, you can show the regulator what a decision was based on with no trouble. Get this right, and transparency and compliance take care of themselves.

This also helps with changes in regulation; for example, the FCA and PRA recently introduced a range of policy changes to increase individual accountability within the insurance sector. With decision management tools, new limitations, rules or guidelines can be accounted for within a decision structure immediately. There’s no need for the disruptive and lengthy internal change traditionally required to reflect new regulations – where it used to take on average eight to 15 months to change all systems, with the right technologies today it can take minutes.

Focusing on decisions rather than data alone allows businesses to build models that identify all relevant decision inputs – including data, business knowledge and results of previous decisions – and thus make better decisions. This leads to efficiency, increased re-use of assets and a more profitable business. Too many companies are drowning in data lakes, unable to realise the value of their investments because the focus was on gathering data. Instead, enterprises should focus on using the most valuable data to inform the most precise and optimal decision for the circumstances. In our experience, businesses that start with the decision get better results, and get them faster. This isn’t always easy to do, but the benefits are worth it.

Cutting time-to-insight is also a game-changer for many in the insurance industry. This is not only the time to deliver a new application, but the time to go from developing an insight through analytics to modifying the analytic models that can put this insight into your production stream. Change management has become a hallmark of solutions as a standard, because that is what is needed in this fast-paced, data-driven business environment.

The codification of industry and solutions specifics – by gathering and recording subject matter on decision processes from an expert perspective, or perspectives – is critical, but also a gap in most companies. Even where subject matter expertise is captured, it is often difficult to infuse it into the systems that drive operational decisions in a business. The latest Object Management Group standard, Decision Model Notation, provides a common approach for modelling decisions, which drastically reduces the time and cost of managing those decisions. It helps different groups within an organisation create a common map of a decision process, which is critical because you can’t systematically improve a decision if everyone involved doesn’t understand it. Properly structured, decision-first modelling gets everyone on the same page. This is the biggest change insurers can take advantage of right now – in order for the best decisions to be made, businesspeople should have the data to inform their decisions and be able to model and deploy it into the operational environment themselves.

Democratisation of analytics is already well underway, but some industries are further along than others. Having an open framework to which you can easily add powerful and advanced analytics is tremendously valuable. Being able to quickly and easily leverage analytics is not only empowering for your business decision-makers, it is the future of business.

Larry Jacobson is a senior consultant at FICO. He can be contacted on +44 (0) 207 403 1333 or by email: larryjacobson@fico.com.

Larry Jacobson

FICO