Round table: Personal, digital, ethical

Pension providers are leveraging big data and artificial intelligence (AI) to better understand scheme members and make their communication strategies more personal. This drive towards hyper-personalisation can drastically improve engagement but also raises new questions in relation to ethics and responsible data usage, hears Muna Abdi

The pensions industry has some time to go before it catches up with TikTok, Facebook and Instagram. Big data and AI are already starting to transform the way providers, advisers and employers talk to staff about their pensions and benefits, but we are only at the beginning of this journey of change, and opportunities and risks lie ahead as new technology is embraced.

DOWNLOAD A PDF OF THE SUPPLEMENT HERE

Delegates at a recent Corporate Adviser/ Aviva round table on harnessing digital to move money mindset discussed the effectiveness of communication strategies in engaging different segments of the population and the need to tailor communications to individual preferences and life stages.

Trust

According to Joe Craig, development lead at Quietroom, it is of paramount importance to set clear objectives and parameters at the beginning of a communications project to define what engagement entails and how it will be measured. This involves establishing measurable outcomes and understanding that achieving better outcomes is long-term and may span years.

He suggested that testing content with a few people can provide insights into a broader understanding without the need for extensive questionnaires for every communication recipient.

He addded that posing wide, open-ended questions promotes trust and motivates businesses to use this strategy.

Craig made a point of highlighting the value of continuously testing and improving communication techniques to reduce the gap between perceived and actual understanding.

He said: “We’re in a position now which we haven’t been in before, where we can ask more questions and get people to just talk and use their own language.”

Craig mentioned the importance of reflecting back the language used by respondents.

He highlighted how people perceive pensions differently based on their stage of life. Many associate pensions with the state pension, with online searc data revealing a strong focus on state pension queries.

He went on to explain that as people approach retirement they tend to learn more about workplace pensions. However, terminology can be confusing, as some prefer the term ‘savings’ instead.

Hyper-personalisation

The concept of hyper-personalisation in communication strategies involves tailoring messages to individual members of the scheme based on various factors such as their age, stage in the scheme, external market conditions, and the objectives of their employer. This can help ensure efficacy and relevance

According to Aviva’s head of workplace client engagement Laura Stewart-Smith, segmentation strategies have evolved to include various factors, including life stages and market segments.

Traditionally, segmentation has been based on demographics such as age points such as mid-life or retirement, as well as market segments defined by what people in a particular demographic might be doing. However, there is a shift towards considering how people feel and whether they take action based on communications received.

She emphasised the importance of measuring audience response and utilising data science to predict and prompt desired actions.

She said: “We are increasing the use of data science and understanding the impact that communications have. We will test A and B types of communications with different audiences, different segments of the market and different points in the journey. We will use data science to build a picture and predict what we think people will do next, and prompt them towards appropriate actions. For example, it might be that somebody is approaching retirement and they’re interacting online with the forecaster a lot. You know these people are looking at their retirement options.”

But there is a limit as to how specific this approach can be. She stressed the need for flexibility and customisation in communication strategies and the danger of assuming homogeneity among demographic groupings.

She said: ”We are looking at how we utilise the data that we hold to bring the right information to people at the right time. It’s a trust thing as well. There can be certain things that we’re never going to have access to unless somebody is prepared to tell us that information. Then if they give it us, we can build on that.”

Big data

According to Stewart-Smith, there will always be data gaps, but the goal is to use the data that is already available to provide people with relevant information at the right times. She said that AI has great potential, but also noted that there isn’t a perfect understanding of how to use it in the world of personal finance at this time.

She acknowledged the efficacy of digital tools and tailored content in inciting action. She also stressed the need for trust when gathering personal data from individuals and the constraints that come with accessing particular data without express agreement.

Emma Douglas, director of workplace savings and retirement at Aviva, noted that efforts are being made to strengthen communication strategies outside of the pension industry by leveraging big data and advanced techniques like speech analytics.

She mentioned how important it is to evaluate communication strategies and use word clouds and other tools to get insights that can be used to improve approaches.

She pointed to the complexities of creating communication strategies tailored to individuals’ diverse life stages, highlighting the challenges in obtaining comprehensive data for pension-related purposes.

Douglas also touched on the effective use of propensity modelling, emphasising its role in predicting consumer behaviour and tailoring communications. This method goes beyond traditional considerations of life stages, offering a comprehensive approach to anticipate individuals’ future actions. She highlighted the predictive aspect of propensity modelling, citing examples like retirement calculators or requests for financial advice as signals of potential changes in behaviour.

She said that these triggers allow companies like Aviva to intervene proactively and offer assistance tailored to individual needs.

Douglas said: ”There’s a lot we can do around this propensity modelling, observing what people are doing. You can overlay life stages on that and add other bits of data.”

Postcode data

Singleton mentioned that she employs postcode data to assist employers and pension schemes in obtaining more comprehensive insights into their membership demographics.

This data aids in tailoring communications, sometimes to specific groups, to align with the insights derived from the information. However, Singleton emphasised that this customisation isn’t personalised to the individual’s address; rather, it’s used to enhance the relevance of the information being sent out based on broader demographic trends.

Additionally, Singleton highlighted that this practice has challenged some schemes’ assumptions about their members’ financial capacities, leading to adjustments in their communication strategies regarding savings and tax benefits.

Ethical considerations

Delegates reflected on the ethical implications of hyper-personalisation and big data and its role within industry practices.

Douglas emphasised it is important not to penalise customers based on data that has been collected. Ethical considerations are paramount in ensuring the fair treatment of customers and maintaining trust.

She said: “There’s a massive ethical question that sits at the heart of this. You have to have the customer’s interests at heart.”

Douglas added that while Aviva hasn’t extensively used data beyond basic information like addresses, it recognises the need for careful consideration of how data is used and the purposes for which it is employed, highlighting the importance of constant monitoring and scrutiny of algorithms to ensure ethical and responsible practices.

Privacy

Singleton noted the increasing integration of personal data into strategies and acknowledged the concerns surrounding privacy and accuracy.

As technology advances, Singleton suggested that assumptions about individuals may become more refined. She pointed out the importance of ensuring accuracy and respecting privacy boundaries and implied that the pensions industry must navigate these complexities cautiously to maintain trust and relevance.

She said: “As technology moves forward, some of those assumptions will become more accurate. It becomes more comfortable for the pensions industry to be able to use some of those assumptions.”

Legal challenge

There are many legal challenges for the pensions industry when it comes to personalising communications. Craig said: “The more you personalise something, the closer you’re getting to hot water in that you can stray over the boundary into regulated advice.”

He said that schemes vary in their risk tolerance regarding the risk that personalised communications could be interpreted as advice, This can influence the level of support provided.

However, he added that new generations are now demanding tailored experiences similar to those they get in other industries and that AI-driven solutions are redefining expectations. This change emphasises how full connectivity—which goes beyond pension data—is necessary to satisfy changing customer needs.

He said: “What you can get with AI is a complete change to people’s expectations of what support looks like. Rather than waiting for your pension scheme to write to you with something that may or may not be segmented, personalised or tailored just for you, you just go and use a chatbot. What you expect based on your experience with every other organisation outside the financial world is that the algorithm you’re talking to knows your information.”

Craig noted that, despite their limits in offering financial advice, there is a growing expectation for chatbots to offer personalised information and guidance. In future is will not be enough to have a chatbot that knows about someone’s pension status; it also needs to take into account other factors of financial wellness, although again, the extent to which having this data results in approach something like a personalised recommendation remains a concern.

Anish Rav, director of pensions policy at Capita Pension Solutions highlighted the importance of informing people about the benefits of sharing comprehensive data, particularly regarding financial wellbeing and the use of AI. He voiced worries that people might need help comprehending why their data is being gathered and how it enhances the user experience.

He recommended that the financial services and pensions sectors focus on showcasing the benefits of data sharing to promote openness and user trust.

Exit mobile version