Other parts of this series:
Stress-testing your personalisation hypothesis
UK insurers have realised the worth of personalising their business practices in order to connect with their customers more deeply. In the previous post in this series I discussed the first step on the path to personalisation, which is to prioritise. Once you have gone through this process – identifying value, seizing initial opportunities and strengthening foundations – it is time to develop and test analytic models and solutions.
The second step: Experiment
With a clear sense of your business’s unique customer-focused goals, you can begin to test your business hypotheses through measurable experiments. You can track the impact of your experiments against key performance indicators (KPIs) and by feeding data into machine learning models.
Data analytics gives us rich insights into how to design the customer experience, and how staff can shift their behaviour to reflect the customer’s evolving needs. In order to succeed in this step, you will need to establish analytics-specific, cloud-based platforms that can tolerate rapid data ingestion, input and analysis. Throughout this experiment, the objective is to balance the customer’s needs with those of the business, resulting in an optimised customer experience.
The personalisation experiment – step by step
Running an effective, customer-focused experiment in your business comprises the following.
- Test hypotheses
The prioritised hypotheses for your business should be tested against measurable and quantifiable outcomes, by establishing randomized test and control experiments.
- Perform experiments
Next, it’s time to perform the experiments. We recommend that you use cross-functional experimentation teams that, typically, include representatives from Pricing, Underwriting, Analytics, Marketing, Distribution and Change functions. This diverse group will undertake the end-to-end design of the experiments, including plan, duration, measures and KPIs (for example, customer lifetime value, retention, and persistency), as well as implement changes to ways of working to introduce the new customer treatment.
- Gather data
Once data has been accumulated from the experiments, it can be analysed and used to identify trends and insights into customer behaviour. Some experiments may take longer than others to secure the statistically significant data volumes required.
- Establish cause and effect
The final step provides the richest outcomes for customers. The data produced from the randomised, controlled experiments is used to establish cause and effect (rather than correlation). This enables your business to create engagement approaches tailored to individual customers.
Experimentation will give you a clear view of customer needs
It’s only through testing your hypotheses and supporting them with robust data that you can forge a path forward in your customer-focused strategy.
If you’d like to read further on the steps your business can take to personalisation, download our report here. We’d love to help you align your business to your customers – you can get in touch with us here.
In the next post in this series, I will discuss how to optimise your insights within your business.