1. The IoT hype
We have all heard about the “Internet of Things” (IoT) as this revolutionary new technology, which will radically change our lives. But is it really such a revolution and will it really have an impact on the Financial Services Industry?
To refresh our memory, the Internet of Things (IoT) refers to any object, which is able to collect data and communicate and share this information (like condition, geolocation…) over the internet. This communication will often occur between 2 objects (i.e. not involving any human), which is often referred to as Machine-to-Machine (M2M) communication. Well-known examples are home thermostats, home security systems, fitness and health monitors, wearables…
This all seems futuristic, but smartphones, tablets, and smartwatches can also be considered IoT devices. More importantly, besides these futuristic visions of IoT, the smartphone will most likely continue to be the center of the connected device universe.
IoT can have serious impacts on both the Business and IT departments of financial institutions. Flexible implementation frameworks and high-performing run-time platforms will become a necessity.
2. Where does it impact us?
In essence, the IoT allows making products and services more personalized, i.e. products and services will be even more centered around the customer’s needs and preferences.
Furthermore, IoT permits companies to capture in real-time (and 24/7) huge amounts of data about the customer, allowing a more intimate understanding of the customer. IoT will therefore certainly mean a revolution for the domain of data analytics, which should turn the vast continuous stream of data into immediate insights and predictions about the customer (what will the customer like / need / do…).
The Financial Services Industry, being a data-driven industry offering intangible products will not have a lot of direct impacts of IoT (in contrast to e.g. the retail industry), but the indirect impacts (i.e. data from IoT devices is indirectly used to improve financial products and services) will be far-reaching.
In the insurance industry, IoT will allow increasing user engagement, which is nowadays low compared to the banking industry. Currently, customers have no incentive to interact with their insurer, and as a result, interactions remain limited to claim requests and yearly premium renewals. IoT leads to a much more intense customer relation, as it allows insurers to develop value-added services on top of their insurance contracts.
3. Use Cases
The different new services and products resulting from IoT in the Financial Services Industry are hard to predict and only limited by the extent of one’s imagination.
Overall, we can classify the use cases according to two axes, i.e.
- The usage of the delivered IoT data:
- Delivery of new innovative products and services (personalization by the IoT delivered data)
- Fine-tuning of Risk Management (e.g. improvement of fraud detection or improved quality checks and follow-up of credit collaterals)
- Improve the sales process of existing products (e.g. identification of cross-selling opportunities, more personalized contextual messages…) and the customer relationship (e.g. churn detection, more accurate customer segmentation…)
- Execute payments, i.e. devices executing automated payments
- Identification and authentication, i.e. use IoT devices to identify and authenticate a person more accurately
- The type of data the IoT device sends or receives:
- Condition or usage of the monitored object or person
- Payment data
- Biometric data
- Customer communication (i.e. human-readable info)
We will structure this article according to the first ax, i.e. the usage of the delivered IoT data.
3.1. Delivery of new innovating products and services
IoT will result in the creation of new innovative, personalized products and services, which would not be possible without IoT.
3.1.1. Home Sensors
Connecting home sensors to a Financial Institution allows providing multiple new innovative services:
- Connect utilities smart meters (water, gas, and electricity) to a bank account, allowing to
- Pay automatically the utilities bills
- Provide a service to the customer to automatically switch between suppliers to get the best deal
- Connect a set of home sensors (e.g. utilities smart meters, smoke and carbon monoxide detectors, fire suppression systems, advanced alarm systems) to an insurance company, allowing the insurer to:
- Improve the protection of insured houses against threats (fire, leak, flood, and theft), thus reducing the risk for insurance claims.
- Provide more personalized insurance pricing. E.g. insurance premiums that increase, if the customer regularly forgets to lock his doors or to turn off his oven.
- Contact a policyholder via his smartphone if a threat is detected and dispatch automatically an emergency response team
- Provide the customer a monitoring view on his house statistics (e.g. online follow-up of his utility consumption)
- Propose to the customer potential improvements to the house, which can reduce the customer’s insurance premium
- Use sensors (like e.g. dampness detector or structural integrity sensors in walls) to determine the condition of the house as loan collateral. This permits banks:
- When an incident occurs, automatically issue a home improvement loan to cover the repair costs and automatically issue a work order to affiliated repairmen. The customer could profit from discounted rates for the repairs, an enhanced warranty, and immediate approval of the associated renovation loan (without any administration).
- Banks could propose a variable interest rate loan, for which the interest rate does not only fluctuate with the central bank interest rate but also depends on the readings of home sensors (indicating how well the customer takes care of his house).
3.1.2. Car Sensors (Telematics)
Where other IoT examples are futuristic, using car sensors (telematics) for providing new financial services and products is already done today by several financial institutions.
These sensors result in a number of interesting services:
- Usage-based auto insurance: measure the driver’s behavior (e.g. kilometers driven, hard brakes, driving in risky areas, driving at night, driving in bad weather conditions, speed…) allowing insurance companies to
- Adapt the insurance pricing accordingly (reward safe drivers and calculate premiums based on actual driven kilometers)
- Minimize insurance fraud, e.g. claimed accidents, which are not detected by the sensors
- Recover cars in case of theft
- Provide the customer with statistics on his driving behavior. When using concepts of gamification (e.g. comparison of driving behavior with other customers or with a list of friends), drivers can be encouraged for safer driving.
- Support customers in case of car breakdown or accident
- Offer fleet management services to SMEs
- Communicate to customers when maintenance is required or warn them of dangerous weather conditions or dangerous drivers in the neighborhood.
- Use car sensors to determine and manage the condition of a car as loan collateral. This allows banks to
- Propose a car loan with a variable interest rate, with the interest rate adapted to the degree the customer takes care of his car
- Remotely disable the car, when a car loan is not reimbursed
- Automatically propose a credit when expensive car repairs are required. Banks could furthermore cooperate with car repair companies to offer discounted rates for their customers.
3.1.3. Personal Health Sensors
Personal Health Sensors are the most invasive sensors when it comes to privacy and monitoring continuously the customer’s activities. These sensors are best suited to get a continuous stream of data about the customer, allowing to get even data about the current mood of the customer. These sensors can therefore also be used for improving the sales effectiveness and customer relationship.
However, when looking at products, which are directly derived from this sensor data, Financial Institutions can think of wearable body sensors measuring health parameters like heart rate, body temperature, blood pressure, movement, calorie burn rate, and alcohol consumption. This would allow insurance companies to personalized life and health insurances:
- Adjust pricing of the insurance based on the health statistics of the customer
- Check if the customer is properly taking his required medication
- Help with safety and care for elderly and assisted living
- Inform policy-holder when a doctor visit is recommended
- Provide the customer a view of his health statistics
- Block car (if also insured at same insurer), if customer’s alcohol consumption was too high and customer intends to drive
3.1.4. Supply Chain Sensors
Supply Chain sensors refer to all sensors to monitor the inventory (e.g. number and type of objects in inventory, condition of goods in a warehouse, detect hazards like mold, toxins…) and the transport of goods (e.g. sensors on shipping containers and transport vehicles).
These sensors are likely to be installed by the manufacturing companies themselves for improving their supply chain efficiency, but the data can also be used by financial institutions to offer new products to SMEs and corporate customers:
- Banks could use this sensor data for multiple purposes:
- Improvement of credit scoring, allowing to provide more personalized interest rates, but also allowing easier approval of customers without a credit history.
- Monitoring of the collaterals associated with corporate credits, but also monitoring of products that are financed by the bank via leasing.
- Allow automatic execution of contractual conditions defined in Trade Finance contracts. This could be limited to checking if goods are physically present at the agreed location, but can also depend on the quality of the goods (as monitored by the sensors).
- Bank’s investment research could also use the sensor data to better predict the future financial results of a company (shareholders could dictate this real-time transparency). E.g. the amount of shipped goods gives a good indication of future revenues.
- For insurers similar opportunities exist:
- Improved pricing of shipping insurances, due to better detection of theft and damage to the shipped goods
- Improved pricing of corporate insurances
Apart from offering these new innovative products, banks and insurers could also use the sensor data to offer new services to their customers, like e.g.
- Provide statistics and real-time dashboards on inventory and transportation of goods
- Automatic notification of any anomalies with transportation of goods
- Assistance in the recovery of stolen goods
3.2. Fine-tuning of Risk Management
IoT also allows for a fine-tuning of risk management algorithms. This can go from managing customer risk (in the context of KYC) over managing credit risk (i.e. risk that customer will not reimburse his loan), insurance risk, and operational risk to managing internal and external fraud risk.
- Customer Risk: using sensor data gives extra data to fine-tune the KYC risk model to determine the customer risk at onboarding and perform the continuous monitoring afterward. E.g. the usage of geolocation data to verify the residence address of the customer.
- Credit Risk: sensors allow banks to get a better understanding of the customer and his risk of not reimbursing a credit, but can also help in better managing the value of collaterals linked to the credits. For example:
- General demographic-based credit models could classify a customer as a high-risk customer for approving a loan, but sensor data about his driving style and the way he manages his home (e.g. is customer economical with his electricity, water, and heating) might give a more positive risk classification to the customer.
- Banks could have a near real-time balance sheet reporting, based on sensors in the warehouses and in the transport vehicles of corporate customers. This allows better managing collaterals of credits, like e.g. the line of credit for working capital that most corporate customers have.
- Customer Insurance Risk: better manage the risk insurers take when insuring a customer. This will typically fit with the insurance-related examples in the above chapter on “Personalized Products”.
- Customer Fraud Detection: IoT data can be used to improve the detection of fraud cases. E.g.
- Credit card fraud detection: use mobile geolocation data in real-time as an additional input into the credit card’s predictive fraud analytics. E.g. real-time match the account holder’s location data with the location of the transaction.
- Insurance fraud detection: sensors allow to identify when incidents occur and what the location of the customer is at that moment (through geolocation). This can reduce significantly customers trying to file fraudulent insurance claims.
- Operational Risk: internally in the financial service company, IoT can also contribute to lowering the operational risk. Different use cases exist here, like e.g. better protection of the buildings, improved monitoring of hardware and networks to avoid service outages…
More advanced monitoring cases can also be expected. E.g. companies could monitor the personal health sensors of their employees for elevated stress levels and patterns of movement. This could allow to better identify internal fraud, but also the risk for burn-out or attrition.
3.3. Supporting Sales and CRM Process
IoT will also support financial service companies in improving their sales and CRM processes. IoT allows identifying the customer’s needs more accurately and instantaneously (even real-time), allowing to perform much more effective marketing and targeted sales. This more intimate knowledge of the customer allows more personalized interactions and therefore improves significantly the relation with the customer.
This paragraph provides some examples in this domain:
- Beacons at the entrance of branches would allow identifying the customer (via his mobile phone) immediately when he enters. This allows the reception employee to
- Welcome the customer by name
- Get an immediate view about the customer on his computer, i.e.
- Full 360° view on the customer’s assets and liabilities
- View on all actions the customer recently did (e.g. where did he look at on his internet banking the last days)
- Overview of any sales opportunities based on customer analytics
- Use the “Personal Health Sensors” and geolocation services to identify the best moment to make a sales call with a customer, i.e. avoid interrupting customer at work, avoid sales contact when the customer is stressed…
- This beacon or geolocation technology can also be used to offer products and services, when a customer arrives at a certain location, e.g.
- When a customer enters a car dealership or a shop for other expensive articles, the bank could alert the consumer on how much financing he is approved for (and offer a preferred interest rate).
- Banks could collaborate with shops via loyalty programs. The real-time location tracking would allow banks to send offers and deals in real-time. The offer could only be applicable if the customer pays in the shop with the bank’s card.
- When a customer is not yet a house owner and is currently located at a house for sale, the bank could send immediately a mortgage offer for the house.
- When the customer arrives in a foreign country, the bank could immediately check if the customer’s credit card is authorized for the country and propose to activate it if it is not. The bank could also offer to temporarily increase the customer’s credit limit, allowing him to pay his hotel bill.
- When the customer arrives at an airport and the customer did not take travel insurance yet, it might be advisable to propose such insurance to the customer.
3.4. Automatic Payment Execution
All previous use cases are all about IoT sending sensor data to the financial institution, which acts upon the data. For automated payments, IoT means that the object takes action itself, i.e. orders one or more products or services and pays for them (without any human interaction).
Typical examples could be:
- A fridge ordering itself products to the super-market
- A car paying itself at the gas station or at the recharging station (in case of an electric car)
- Smart assistant on a mobile phone ordering itself airline or movie tickets
- A blown light bulb automatically ordering a replacement
Supporting these use cases will still require a considerable evolution in the payments industry, since the number of payments would drastically increase (with lower payment amounts), meaning that the costs of payment should also decrease accordingly (to avoid customers and banks having to pay increased costs). Furthermore, IoT devices need to be correctly linked to a user or even more precisely to a bank account.
Often the use of blockchain is put forward as a solution to these issues.
3.5. Identification and Authentication
IoT can also help to improve the identification and authentication of customers. In an ever-increasing digital world and ever-growing concerns of digital security, IoT can bring a good solution to this concern.
Today identification and authentication are still mainly done by the combination of “something you own”, i.e. typically a card (bank card or identity card) and “something you know”, i.e. typically a PIN code or password. Today this type of security is no longer sufficient.
IoT can bring a solution through different types of validations, based on “something you are” (i.e. biometrics checks).
- Fingerprint check on your mobile phone
- Iris scanner or face recognition via your mobile phone
- Verifying the way, you hold your mobile phone
- Connecting with wearable technology and comparing with your profile (your typical heart rate, blood pressure, body temperature…)
- Checking your last-known locations with the location where the request is made
- Validate certain behavioral patterns derived from IoT with the current behavior
4. Required IT Capabilities
Banks and insurers will dramatically have to change their IT infrastructure and application landscape to deal with IoT.
The main game changers for IT will be:
- The capture and processing of data and all interactions with the customer should happen in real-time and should be 24/7 available. In an industry, which is still largely batch-driven, with significant windows of unavailability (for batches, maintenance…), this is a real challenge.
- IoT will deliver a continuous stream of large amounts of data. This will exponentially increase the volumes of data financial institutions have to deal with. A redesign of the data architecture will be required to deal with this (bandwidth, disk storage, compute power…), but also to filter, enrich, and process (e.g. real-time customer analytics) this data (near) real-time. Potentially a migration to cloud solutions (companies specialized in coping with these volumes) will be the only logical step.
- With IoTs collecting highly personal data (e.g. about customer’s health), security, safety, and confidentiality are essential. This data should be protected from both the outside, as from the inside (you do not want your banker to have a view on your body temperature of the last week). On the other hand, you do want your banker to be able to explain why a product or service is suddenly becoming more expensive (because of certain sensor data).
Furthermore, IT should find a solution to link a customer and/or account to a device. For devices only used by 1 user (e.g. smartphones and personal health sensors), this is relatively easy, but for devices shared by many users (e.g. thermostat or car telematics), this can become more complex.
- The IT infrastructure should support different devices, with most likely different communication protocols. Furthermore, the IoT devices will regularly evolve (e.g. deliver new types of data).
A highly flexible architecture with very low development times (supported by different framework tools) is therefore essential.
Gradually IoT devices and the data collected by them will take a more prominent role in the services and products offered by financial institutions. Today most of the usage of IoT in the financial services industry is still in the experimental phase, but with an exponential increase in usage, this can rapidly change.
Banks and insurers should therefore act now in changing their application architecture and more specifically their data architecture, to be able to support these future use cases. The typical evolutions to DevOps, Agile, Cloud, and microservices can be an excellent preparation for this IoT (r)evolution.