9 September 2015
CHECK AGAINST DELIVERY
Thank you Rory. Well, good morning. And before I start, I would like to thank you all for coming today and to thank also the team at the ABI for organising what looks set to be a fascinating conference.
Data. Well what can I share with you about data? We’re insurance and financial services companies after all, and have surely used data since time began? Data is our raw material. Our collection and analysis of data is the competitive advantage that we each seek to build against each other, and also the barrier to entry that protects our market from those who would pretend to enter. Or is it?
Well it was not always thus. When our forefathers began offering marine insurance, for example – they used publicly available data on ship losses and relied as much on the law of averages in insuring enough ships to cover their likely losses. But over time, insurers have captured masses of loss data – car crashes, stolen cars, break ins, property fires, floods, premature death and illness – so that we are uniquely equipped to leverage that data to price risk.
And whilst for marine and aviation insurance some of the claims frequency data is still publicly available if you’re prepared to search hard enough, only insurers have captured the vast majority of risk and loss data in the market today. And that is our competitive advantage.
So why have a conference on data? Surely we are the experts? Well we’re all here today because we recognise that we’re amidst a paradigm shift and personally I believe that how we respond to the opportunity this brings, will determine the future of our industry.
Many of you will have heard of Moore’s law, which in simple terms says that processing speeds will double every two years. First conceived in 1970, we have now seen 45 years of experience – or 22 iterations of the doubling effect - and so far the law has proven correct.
It is worth illustrating the phenomenal impact of Moore’s Law, and the best illustration I have seen uses a chessboard.
I take a normal chessboard and I make an offer to my boss. I say that I will cut him a deal and for this week I will work for one penny, and at the end of the week, he should put that penny on the first square of the chess board. But next week I want double that, so just two pence, which he should put onto the second square, and the week after I want double that, or just four pence on the third square, and so on. I offer that once the board is full I will work for free until I retire. Should he accept my generous offer?
Well not if he has a calculator – because by the time the chessboard is full, I will have earned 1 septendecillion pounds; that’s 17 zeros. Give it a try when you get back to the office!
And that is the exponential scaling factor of Moore’s law. And the really scary thing, is that after 22 iterations we are not yet even half way across the chessboard – with processors 2 million times faster than in 1970.
But why is Moore’s law relevant to a conference on data? Well because the rapid acceleration of processing speeds and the miniaturisation of devices this permits, has brought about the information age. And if anything the proliferation of data is accelerating even faster than Moore’s Law. It is said that 90% of all the data in the world was created in the last two years.
And that acceleration will continue with the development of connected devices. Ten years ago there were already 500 million connected devices – mostly phones; today there are 8 billion and analysts forecast there will be 50 billion by 2020, but incredibly, 1 trillion connected devices by 2030. Once again following Moore’s Law of doubling every two years.
So within the timeframe of the careers of most of us in this room, almost every interaction our customer has with the outside world will be captured by a connected device.
That is already creating an enormous volume of data, which is only going to grow exponentially, and which will provide phenomenal insight into the risks society face as they go about their day to day lives.
And therein lies the opportunity before us, but also the profound threat.
How we behave during the data revolution may well determine the future of our industry. We must remember that access to, and analysis of data will not remain unique to insurers – others will seek to harvest and commercialise the data from connected devices – car manufacturers, utility companies, companies that haven’t even been formed yet. We’ll no longer have an automatic right to know more about the risk profile of our customer than others, or potentially even the customer themselves, potentially losing the competitive advantage and barriers to entry we have today from this ‘asymmetry’ in data knowledge.
So what should we do about it?
Well let’s first think about the term ‘Big Data’. We’re all using it, but I wonder how many of our customers really understand what it means? Try googling ‘Big Data’ and then ask yourself how many of your customers will understand the concept of ‘structured’ and ‘unstructured’ data?
Remember, a recent Accenture report said that less than a third of customers trust their insurance company. And in a world of mis-trust, ‘Big Data’ sounds very much like ‘Big Brother’ to me, and customers need to understand why we want to access their data, what we are going to do with it, and, most importantly, what’s in it for them?
And this last point is key – what is in it for them? I have already spoken out about the risk that insurers use data to identify and extract all the great risks, at a great price, but with the consequence that we create an uninsurable class in society – those whose data tells us they are significantly above average risk and should be charged their true, unaffordable, risk price, a price that is not cross-subsidised by the better risks in the portfolio – most probably weighted to the poor and disadvantaged - the same class that our society prides itself in seeking to protect.
In our desire to derive a more precise risk profile of each individual customer from the data they generate, we must not lose sight of our responsibility to mutualise risks so that insurance remains affordable and accessible for all.
If we don’t, then very quickly we shall find we have lost our right to access that data – either through customer behaviour, or through Government or regulatory intervention.
So before we all get carried away by the opportunity of Big Data, we need to think about how we talk to our customers about data, and sit back and agree our moral code.
I think that most people, when they understand the data trail their lives are creating, will understand the benefits of sharing that data with us.
Recent Deloitte research, for example, suggests that four out of five customers are willing to share their personal information if the benefit is sufficiently compelling to do so – such as better coverage, lower premiums or a more personalised product offering.
But to be trusted do so, insurers must make clear and transparent promises about how that data will be used.
Take for example the report the ABI has published today – “How Data Makes Insurance Work Better for You”.
I encourage you all to read it, because it is a great overview of the Big Data context for our industry – indeed I wish I had seen it before today because it would have made writing this speech a great deal easier!
Within that report, the ABI have reinforced our commitment to the duties we have in how we treat data:
- First - We recognise that your data is amongst your most personal and valuable possessions, and it is always treated as such – safely, securely and confidentially;
- Second - We do not collect any more of your data than absolutely necessary, and only store that which is essential to achieving one or more of the following uses:
- Providing you with a more personalised product, which accurately reflects your individual risk and circumstances;
- Understanding and improving our customer service;
- Improving our business in a way which ultimately benefits customers.
- Finally - We only use data which has been collected in the knowledge that it will serve one of the uses above.
Let’s look at the first promise – that we’ll handle your data safely, securely and confidentially.
Now the general insurers amongst us will see the emergence of Big Data, and the ever increasing exposure to hacking, as a real business opportunity. Businesses need to protect themselves against the financial consequences of a major hacking incident; which in extremis could destroy a business. You might, for example, expect that the recent hacking of Ashley Maddison – ‘Life is short. Have an affair’- would have done untold damage to the business; ironic perhaps that they have confirmed that they continue to attract new members – though I rather suspect many are the ‘bot’ created female members which they use to attract men to the site.
Companies will need insurance to protect their business from hacking, so ‘Big Data’ presents general insurers with a significant new market opportunity. Not an easy one to price of course - in the good old days a bank robber had to walk into a bank with a gun and run out with as much cash as they could carry whilst hoping they are not shot in the process. Today, millions can be stolen from the safety of your bedroom.
So we have the opportunity to insure the data of others, but of course the most fundamental requirement of us as insurers is to protect the customer data we keep ourselves – especially as that data grows and its manipulation becomes a wider, and more dispersed, part of business activity. The information that we hold is already attractive and valuable to hackers, and that will only grow. Today I’m sure we’ve all excellent IT security over our financial processes, but is every customer database afforded the same protection? I hope so as it’s clearly information which is either valuable or embarrassing when in the wrong hands – for example, when Anthem’s medical database in the US was hacked, the medical data of between 10 and 20 million people was made available to those who would abuse it – either for fun, or for gain.
So we must work harder to protect the data we keep, or we will undermine the trust of our customers, and ultimately lose the privilege to access that data at all.
It will be very interesting to hear what the Information Commissioner says later today.
But once we have protected our data, our primary goal must then be to use it to enhance the experience of our customers. The ABI statement promises that we’ll collect only data that is absolutely necessary, but I think we’ll easily demonstrate that the more data that we collect, the more relevant we can be for our customers. But we must demonstrate that value to the customer.
That might be by:
- A deeper understanding of the needs of our customers so that our offers are more timely and relevant.
- A deeper understanding of the risk profile of our customers so that we offer the right level of cover at the right price that rewards their behaviours.
- A deeper understanding of the value to us of our customers so that we reward their loyalty.
- It will help general insurers to start managing a claim before the customer even realises they have one; thanks to the reports from connected devices. Form filling could become a thing of the past; reducing our costs, and the customers’ premiums whilst making their lives much easier.
- But perhaps also so that we start to anticipate risks for our customers and help them to better manage those risks.
For example, even today many of us use telematics data to suggest how customers can become safer drivers. These ‘Asbo’ bracelets many of us now wear gather enormous volumes of data on our sleep, our exercise, and our fitness level – data which is then used to help us change lifestyles for the better.
This goes back to my earlier point about earning the trust of our customer. With good use of data, at one extreme, we can take the place of a trusted lifestyle interpreter and adviser – evolving our positioning from paying out when risks materialise, to helping encourage customers to avoid those risks materialising in the first place.
Data represents a fantastic opportunity to deepen our relationship with customers, and thereby to earn their trust. As I have said many times before, that insurers are trusted less than banks, or worse, less than estate agents, is frankly humiliating and we must collectively and individually do more to address this –good use of data should help that. Correspondingly, bad use of data will undermine trust forever!
So to summarise – Big Data is already Big, but it’s going to get much Bigger over the next decade. Some refer to data as the new oil. I think they’re right, and managed well, and to the advantage of the customer, it will certainly fuel a revolution in our industry – but manage it badly, and we’ll have more than the environmentalists to deal with; the very survival of our business model will be at risk.
Thank you all very much for taking the time to listen to me today. I’ll now hand you back to Rory, your host for the day.