June 2016, London, United Kingdom
Digital and big data collide in a brand new connected world for insurers
Price Sensitivity and the Line between Creepy and Insightful
The insurance industry is undoubtedly transforming the way it presents itself to the consumer: omni-channel engagement, digital interaction, ultimate price transparency, personalized offers and lately the data driven discounts made available by the emergence of connected devices such as the telematics black box.
These advances have incontestably benefited the insurance industry. Advancing the existing data sets has allowed further developments in statistical modelling of big data that is now widely used not only on the actuarial side, but also in underwriting, pricing, claims, fraud and marketing.
The more consumer information an insurer has, the more accurate risk modelling can be. This results in more suitable price, increased fraud prevention, more tailored coverage and personalized offerings for customers. Even though some consumers may be aware of the benefits that they will eventually receive from providing insurance companies with more personal information, some people are still cautious when it comes to sharing their personal data.
A number of surveys detailing the extent to which consumers are willing to share their data are available, and will be referenced throughout this paper. When we will be looking at our own survey results, where we asked 500 consumers of a wide demographic range their preferred ways of engaging with insurers, we found the extent they would be willing to share their data and their general perception of the industry.
According to the leading consultancy McKinsey & Company, there has been a gradual shift in the value that carriers can provide to customers during many activities traditionally performed by local agents. FC Business Intelligence’s findings from interviews with US insurers found that 50% believe that disintermediation will continue and identified commoditization of insurance as one of the 3 main challenges, whereas European insurers identified low premiums that were accelerated through the emergence of aggregators to be one of the main issues.
There are obvious variations in the channels used and preferred across different products, with the lines of business that are more bespoke and have less potential for commoditization (i.e. life and health) facing more intermediation than auto, travel or phone insurance. According to the McKinsey Report, auto insurance is the most disruptive to the traditional agent.
So, we asked modern consumers which channels they would prefer to use. 72% of European consumers identified online through website to be a method they prefer, 38% selected aggregators and 23% the agent/broker channel.
In North America, the broker channel was significantly more popular with 43% selecting it amongst their favourite; 57% choosing online with 29% saying that they would prefer an aggregator. Even though 23% of millennials use an agent or a broker, only 16% identify it as a preferred insurance channel (making it the 4th most popular choice after online, aggregator, and affiliate partner channels). It also seems that millennials much prefer affiliate channels than other generations.
61% of Gen X and Baby Boomers used an agent or a broker with 50% selecting that channel as one of their most preferred. However, online still remained the most popular and preferred channel. The vast majority of respondents selected multiple channels - meaning that the omni-channel engagement trend is continuing.
What makes people buy insurance?
One of the main reasons why aggregators have gained such popularity amongst consumers was increased price transparency.
Insurers, after realizing that most consumers do not look past the top 5 cheap-est offers, entered a form of “low premium war”, that even though beneficial for consumers, as all Bertrand-style price competitions are, could have severe adverse effects on their own profit margins.
But, is price really all consumers care about? What about choosing the policy coverage that is best suited for your needs? Does brand or recommendation still mean anything to the modern consumer?
We asked people to rate the importance of each factor when purchasing insurance.
Policy coverage scored the highest with a weighted average of 8.74, and price was slightly behind with a mean of 8.71. Both variables scored a median of 10 (meaning the majority of people asked selected them as being the main influencers of the decision). 79% of people identified insurance being too expensive as the main reason that would change their decision from yes to no.
Unsurprisingly, younger consumers placed more importance on price, ranging from 9.1 for the 15-24 segment to 8.4 for the 55-64 group. However, insurers must continue their efforts to improve and maintain a high quality of customer service, as this has scored 7.82.
Brand averaged 5.41 and knowing someone that uses a certain provider also scored a relatively low 5.3.
So, we’ve seen that for most people an ideal insurance policy would be one that would cover all a consumer’s needs and be affordable.
From an insurer’s perspective, this is now becoming possible to achieve with the advent of big data, first born of increased use of technology in our daily lives and now exaggerated by the introduction of connected devices.
But, are customers really willing to share their data? And, if so, where is the line?
Canadian data and loyalty analytics firm Aimia, in their Global Loyalty Lens report, which features results from 20,000 consumers across 11 countries, found that 80% of consumers would share their names, email addresses and nationalities, while 70% would share their dates of birth, hobbies and occupations.
However, Digital Catapult, a national centre which aims to advance the UK’s best digital ideas, found that 60% of consumers admit they are uncomfortable about sharing their personal information with any organisation, with 14% flatly refusing to share any personal data at all.
The CMA found widespread concerns about how such data is collected, incomplete awareness amongst consumers and hence a fragile level of trust – all of which has the potential to undermine benefits in the future.
Future Foundation presents an idea that consumer preferences regarding data sharing are not homogenous. They identify 3 archetypes: pragmatists – who consider what’s worth sharing case-by-case; fundamentalists – who are opposed to sharing online and not concerned – who frankly don’t care either way. Today, slightly more people take a pragmatic view, the proportion increasing slightly, from 53% to 54% of the total. There are fewer fundamentalists, this proportion dropping from 31% to 24% while those not concerned grew from 16% to 22%. Clearly consumers are warming to the idea of data sharing. A PWC survey reveals that the majority of consumers (73%) were willing to share personal information, depending on the benefits they will get in return.
So, to what extent do these findings apply to the insurance industry? Are customers willing to share the same data with their insurer as they are with their favourite supermarket, chosen social media site or search engine? We tried to understand what constitutes a significant enough benefit for customers to reveal more about themselves and to look into the reasons why some may be reserved about doing so.
We asked which type of data consumers would and would not want to share with their insurers.
The results for basic information (age, gender, marital status) were in-line with the results of the previous surveys, with 81.5% saying that they wouldn’t mind sharing this data with insurance companies. The majority would also not mind sharing their employment information (60 %).
Amongst the type of data that consumers would be least willing to share would be social calendar, with 71% not willing to disclose that information, 64% not willing to share their social media information, 63% not willing to share data on their purchasing behaviours and 59% not willing to share their location. There also seems to be a small increase in the willingness to share data from all categories in the younger customer segments. Furthermore, the data trends support the existence of distinct groups, with an evidence of around 25% selecting 5 or more out of the 7 types of data included in the question, which would suggest that they belong to the fundamentalist group.
We asked people for the main reasons why they may be reserved to share their personal data with insurance providers and 79% of respondents claimed that it was fear that it will be shared with 3rd parties. Data from the latest Information Commission Office (ICO) Annual Track survey concluded that 85% of people are concerned about how their details are passed on to other bodies.7 This is also in line with Altimeter’s results, where they found that 78% are concerned about their data being sold.
51% of people admitted that there is no particular reason for their reservations - they just simply want to preserve their privacy, and there is a significant group of people (43%) that find overly personalised messages creepy.
In their comments people also exaggerated the aspect of trust:
“Because if the insurer can track data back to me, they may want to supervise me. If I say I’m sick they should simply believe me not go on my Instagram to see how I’m dancing at a party. Insurers will have to move very cautiously here and earn a lot of trust before they can ask for these things. If you don’t have trust just asking is crossing the line! (Consumer, UK, aged 25-34)”
So what is the industry response to the lack of trust?
Insurers released a statement on the uses of customer data, which highlights that the industry recognizes that customer data is amongst their most personal and valuable possessions. The statement explains that insurers always treat data as such - safely, securely and confidentially; they do not collect any more data than absolutely necessary, and only store that which is essential to achieving one or more of the following uses: providing a more personalized product or price, understanding and improving customer service, improving business to benefit customers and only using data which has been collected in the knowledge that it will serve one of these uses.
Furthermore, the FCA plans to look into how firms use so-called Big Data: details of customers’ lives, from the way they drive to where they shop. The FCA says: ‘We will conduct a market study to investigate how insurance firms use Big Data, such as web analytics and behavioural data tools, including the increasing use of social media, as well as other unconventional data sources. Even though it seems that these are steps towards better transparency regarding data use, insurers still have a long way to go.
However, as the PWC survey already established - customers are willing to share their personal data for evident benefits. Yet, it is not always straightforward to measure the type of data customers may be willing to reveal for a certain benefit. We came up with 3 situations that were asking consumers whether they would share a certain type of data and if so, what type of benefit they would want.
An Insurer offers a customer a discount on health insurance, on the condition that they wear a Fitbit or any other wearable device and share their activity data with them.
33% of those surveyed said that they would not consent to sharing data from their Fitbit at all (whereas when the same sample was asked whether they would mind an insurer using their lifestyle data from a wearable device without presenting any hints about the benefits, 46% said that they would not share that type of data).
We also asked people what discount they would need off their health insurance to consent to something like this.
Responses ranged from 0 to 100 %, with a quarter of respondents specifying that they would require the device to be included as well.
The average discount requested was 37.5%.
The comments range from people being very positive about the idea and saying that they wouldn’t need a discount.
“I don’t think I’d need a discount if it helped me with fitness, depending on what information the insurer gave back to me” (Consumer, UK, 35-44).
“I wouldn’t expect a huge discount if I also got something that was useful to me as well - like a Fitbit. If I got nothing that was of value to me and my data was still collected, I’d except maybe 10% - maybe more or less depending on the kind of data they’re obtaining and the type of insurance” (Consumer, Asia-Pacific, 15-24).
Some people were very sceptical, saying that insurers should actually pay consumers for their data.
“This data is too private and is personal property. Knowing how valuable this kind of data is, insurance should be free, or the insurer should pay money to the individuals it wants to insure” (Consumer, UK, 25-34).
“No, because that flips the “moral hazard” situation on its back. Currently, I could be as unhealthy as I like but insurers can only check at specific points in time and if the situation becomes worse they are none the wiser. I don’t want them to start building models that enable them to price in my habits (heavy nights out, the odd cigarette, greasy foods at 4 o’clock in the morning, erratic sleeping patterns, lots of coffee and sweets, yo-yo dieting...). Even if I was on the healthier side of the scale, insurance isn’t that expensive so even a substantial discount wouldn’t make me care. Maybe I would see more value in a discount when I’m older or if I move to the US where private health insurance is a must-have and expensive”(Consumer, UK, 25-34).
An insurer offers the customer an exclusive app that helps them to locate an empty parking space when they arrive at a destination. To use this app, the customer must install a sensor in their car and share data on their driving behaviour with the insurer.
Here we presented a benefit that would be of a lower value that is non-monetary.
46% would not consent to this, highlighting that the exchange is not fair in their eyes.
“No as locating a parking space is not vitally important and it’s part of the fun of driving. Also I would like to minimize the amount of personal data shared with an insurer to the bear minimum needed for the insurance to be valid” (Consumer, North America, 55-64).
Furthermore, a number of consumers admitted not being comfortable with install¬ing an extra sensor as it would reveal their less than exemplary driving behaviour.
“No, because then they would know I usually speed and hence wouldn’t have to cover me if I did happen to have a crash” (Consumer, North America, 55-64).
“No, mostly because it will catch you speeding, which it’s now almost impossible not to do with the new 20 mph speed restriction zones” (Consumer, 18-24, UK).
“Hard to believe this app will be precise, different cars need different space in parking places and because my driving behaviour is terrible, I do not want to share this data with any insurers” (Consumer, EU, 55-64).
7% of people said that they may consider sharing their data depending on implications, and 41% of people said that they would be happy to share that information and once again in the comments they revealed that the reason for that would be that they consider themselves to be good drivers.
For the next situation, we presented consumers with a smart home scenario, i.e. where all their devices, appliances and home infrastructure is connected, and can be observed and controlled remotely.
We then asked whether consumers would be comfortable with their insurers monitoring information on their smart home and informing them on any potential hazards, like a pipe that is about to leak.
The majority of people (58%) said they would consent to this, some providing a condition that only the home equipment should be monitored rather than personal movements within the house.
“Monitor my pipes, just don’t monitor me!”(Consumer, UK, 25-34)
22% said they would rather find that information from the appropriate (gas/ electricity/water etc.) provider, and 21% said that this sounds invasive.
“That sounds invasive and rather creepy, but then I can see the benefit - just not too sure how comfortable I would feel with this” (Consumer, UK, 25-34).
So, we’ve seen that consumers would share their data if they perceived the value exchange to be fair. However, when customers have a choice on whether to share their data or not, the moral hazard problem comes back into place where only those that know that they are in the lower risk category will opt in for the data sharing, as they know that it will benefit them.
Insurers must be more transparent about the use of customer data, ensure that it will not be sold to 3rd parties and clearly outline the uses and benefits to the consumer. The benefits themselves should be tangible and customized.
Insurers need to increase the trust that consumers have in them and the way they perceive them.
48% of consumers think that insurance industry is lagging behind technology; 19% think that it is one of the most old-fashioned industries out there; 25% think that it is in line with technology and only 8% perceive it to be forward looking.
Insurance is clearly not appearing to be a tech leader in the eyes of consumer! Or maybe, it is not given a chance to - there are a number of exciting features that innovative insurers are coming up with and most insurers are realizing the importance of putting the consumer at the top.
So, hopefully that will change the way that the industry is currently perceived, yet until then insurers will have to work hard to win the price-aware and privacycautious modern consumer.