This article was a runner up in the 2020 PMI Student Essay competition.
It’s Friday night, and I’m sat on my sofa deciding where to splurge some cash on a takeaway. I take out my phone and contemplate the dozens of options laid out before me: Indian, Chinese, kebab, pizza, you name it. I seamlessly scroll through looking for something that catches my eye. Prices, delivery times and special offers are all right there. Two clicks and 20 minutes later, dinner is served. Ten minutes after that, I’m in a blissful food coma.
Most of us think this is a pretty unremarkable transaction. We take for granted the technical and logistical complexities of delivering this type of service; we simply expect it to happen. But the technical hurdles are actually very significant. And have you ever considered how similar the service is to what we’re trying to achieve with the pensions dashboard? Let me explain.
My go-to is Deliveroo. The user interface is the best, it’s simplest to use and, based on my considerable, detailed research, it’s the most reliable and accurate on delivery times. Deliveroo, a business started in 2013, is now, a mere six years later, valued at over $2bn. The success of Deliveroo and others of its kind is founded on a universal truth – we are lazy, and we enthusiastically embrace technology that makes life easier. Deliveroo is a successful business – but a business that doesn’t supply the core product. It’s more than a courier, website or price comparison site, but less than a consultant or adviser telling you what’s best.
Deliveroo, just like the pensions dashboard, doesn’t supply the end-user product. Instead, it knits together a complex network of different providers to provide the consumer with a single consolidated view of all options. In fact, what Deliveroo manages to do is far more complicated than that and it’s surprisingly similar to the challenges we’re facing implementing the pensions dashboard.
One of the major hurdles that the pensions dashboard faces is how to capture data from many different pension providers. Each restaurant, chippy and deli that uses Deliveroo has its own systems and sources of data, but each must submit data to Deliveroo in the same format so that the user’s experience isn’t an incomprehensible mishmash of content. It’s the same for the pensions dashboard. Okay, I admit that pensions data is a little more complex than the price list for chop suey and pad Thai, but restaurants also have to submit dynamic data changes on stock availability, special offers and live capacity. And, while pensions data might be a little more complicated, you would expect your pension provider to be more technologically advanced than your local Chinese takeaway.
All in one place
In a world before Deliveroo, you had two options for ordering your Friday night beano:
- Root through a packed man-drawer to find an out-of-date paper menu (probably next to your last benefit statement). Try to find out if the restaurant is still open (your pension provider) and what options are available (can you still transfer?). Get out a scrappy piece of paper, note down what everyone wants (your policy details) and engage in some painful telephone exchange to get the information communicated (through the helpline).
- Randomly go from website to website, checking whether the information you want is there, checking opening times and menus and eventually going through some laboured check-out process. Again, sound familiar? This is how most people find details about their pension and attempt to communicate with their providers.
Compare, contrast, contract
The very point of the dashboard is data consolidation and presentation – getting everything in one place so that the customer can make fast decisions about what they want to do. It’s a simple user interface that gives the consumer enough information to make a decision. And a lesson for us on the next evolution of the dashboard is to get transactions happening easily, quickly and with zero friction.
Predicting the future
If you were to boil it down, the main problem facing the pensions industry is this: it’s impossible to predict the future. If we could, then members, providers and sponsors would have a significantly better experience and many of the industry’s ills would be resolved. If we could quickly adapt to a changing environment, we could help members plan for better futures. Charges, returns and contributions are all complex variables that can materially change long-term financial outlook. When drawn together into a single view, accommodating for variations and the impact they have as comparators becomes remarkably complex.
This is where you might think the pensions industry wins out and beats Deliveroo hands down. Our complex models and experience of predicting the future surely can’t be bettered by food couriers? But your assumption is wrong. Deliveroo is far better at predicting the future than we are, and far better at explaining the changing variables to the consumer.
The technology that sits behind Deliveroo is called ‘Frank’; it continually has to calculate and match the best combination of riders with customer orders. It also has to predict the level of future demand, and the location and concentration of requirements. Apart from macro demand changes such as public holidays, demand in each city and district for Deliveroo’s services varies depending on local events such as sports broadcasts, the weather, public holidays and paydays. Deliveroo uses sophisticated machine learning to predict these demands and to cater for them by securing more riders in the right places. Impacts to customers are fed back instantly through variable delivery times.
So if we’ve managed to harness big data sufficiently to predict human behaviour relating to something we perceive as totally personal and unique to us – our stomachs – surely we can do something similar for pension savings.
The centrepiece of Deliveroo’s success has been its ability to commoditise the linked-up market. It adds value to providers by providing tools and insight that make their operations more efficient. And for customers, it provides a service that streamlines their experience, puts all the relevant data at their fingertips and ensures they get what they want, when they want it.
All this convenience for a relatively small cost is great, but Deliveroo is successful because of its near-perfect end-user experience. It knits together complex sets of data from a huge number of different sources and keeps it up to date in real-time – all with very minimal overhead from each data source. Instead of just meeting the functional aim of connecting A with B, it has developed a solution that uses a mass of consumer data to build models that predict the future, for the benefit of its customers. It’s a challenge that we in the pensions industry face every day.
The pensions dashboard is a technology challenge, not a pensions one per se. The problem of how to handle multiple data sources with complex variables is not unique. If we really want to learn the best way to meet this challenge, we need to look beyond our small frames of reference, beyond financial services and into those technology solutions that have had a hugely disruptive effect on their markets, no matter what they do. And what better place to start than a business that has had a revolutionising impact on improving my Friday nights?
Peter Laurence, Senior Pensions Associate