SPONSORED CONTENT
For decades we’ve grappled with the problem of getting people to engage with their pensions. Success has eluded the industry somewhat.
We’ve tried strategies from Plain English pledges, webinars, pension surgeries, interactive PDFs, videos, pension buses, projection tools, finger-wagging, carrots, sticks, even pension raps from Big Zuu. But has this made a significant difference?
For your average pension communication expert the Holy Grail has been ‘engagement’. Often measured in quantitative terms – click through rates, login stats, seminar attendees and webinar registrations. If people are hearing us talk, or logging into our Apps then we must be doing things right. Right?
Engagement rates for members with pension portals have loitered around the 12 to 15 per cent mark for the best part of 15 years. What do we want? Do we want 60 per cent of our members visiting their pension savings 20 times a month? Obviously not. Pointing to ‘high’ engagement levels doesn’t really feel like the right measure any more.
Contrived engagement is not the same as purposeful engagement. Maybe the goal isn’t to talk to everyone. Maybe the goal is to be there when people want to talk to us. That’s a very different definition of success.
Right place, right time
What we want is for our customers to be able to have personal conversations with us and to trust what we tell them. We want them to know where we are, and to come to us when they need help, or want to do something. This is where AI, Large Language Models (LLMs) and technology can make a real difference.
We know that a communication that feels personal and relevant is one that is likely to engage someone the most, and likely to build that most elusive commodity – trust. Until recently we’ve had little choice but to navigate our own way through tonnes of data and information. As is often the case in the digital era, the problem is not a lack of information, it’s too much.
But LLM agents can sift, slice and select like never before. You ask it questions and, if it’s a good AI agent, you’ll get answers back that are on the money. This contrasts with traditional chatbots, which use pre-defined rules to provide an approximation of an answer based on a set script and some ‘if-this, then-that’ logic. It’s the difference between your Sat Nav taking you to within a mile of your destination and taking you to the front door. Close is not close enough.
But all LLMs are not the same. The super AI LLMs, like Chat GPT, literally scrape every piece of content from the web. In some senses, their reference points, are the sum of all human knowledge. But even Chat GPT is not yet a specialist on every conceivable subject.
But there are alternative AI agents that adapt the large language model to reference a body of specialist content. These models are ‘trained’ to become subject matter experts in a particular domain of expertise. The better they get, the more useful they become and the more they can be trusted. The value isn’t in the tech so much as the quality of content, training and guardrails.
So, what about being in the right place at the right time? Maybe we need to learn from Jeff Bezos. Let’s stop trying to get people to come to us and, instead, we go to them.
So where are they? One thing’s for sure, they’re not clamouring to log in to pension websites 20 times a month – and who can blame them? No, we’re all on Facebook Messenger, Snapchat, YouTube, TikTok, TEAMS, WhatsApp, Google Assistant and a whole host of others.
Embedded finance
AI LLM models that can plug-in to these platforms and serve-up their own specialities, expertise, customer journeys and prompts will revolutionise industries like workplace pensions. They can remove much of the friction of the search for answers that we experience today. For the first time, we can be on hand with pertinent, useful support exactly when people need it and wherever they happen to be at the time. Download your workplace pension plug-in, link it to your favourite private messaging service, and you’ll never need to remember your memorable phrase again for the App you registered for back in 2015. AI and LLMs can realise the dream of embedded finance and, in our case, embedded pensions.
Give it two years and your pension provider may have finally given up the incessant email spamming inviting you to login to your pension App. Instead, you’ll have a plug-in that means you’re only ever talking through WhatsApp, or wherever else you like to go. A question occurs to you like, ‘what’s my pension worth?’ You get an immediate answer and then you’re asked if you want to know the value of your other pensions. You say ‘yes please’, and it tells you. You ask what those pensions will give you when you stop work and it tells you. It then asks if you want to think about collecting them into one pot and you say ‘yes please’. And it gives you an application widget there and then that takes you three minutes to complete. And if you want to be doing it at 22.45 at night when the kids are asleep, then that’s OK too. Imagine that! That’s what we’re doing.