The Limited Times

Now you can see non-English news...

Artificial intelligence in insurance: Mr. Kaiser has had its day

2021-01-28T06:25:49.332Z


The insurance industry processes vast amounts of data and is particularly suitable for applications of artificial intelligence. Chatbots, telematics tariffs, damage assessment and "dark processing" by AI should help customers and providers - an overview.


Chatbot instead of consultant in the future?

Only the older television viewers remember the Hamburg Mannheimer advertising icon, Mr. Kaiser (Image: Archive)

Photo: Youtube / Hamburg Mannheimer

"Hello Lord, Emperor!"

Many still know him, the smiling insurance agent.

As a TV advertising icon for Hamburg Mannheimer, he was always close to customers for 35 years.

Insurance brand and mime have been history since 2009.

However, in the near future, applications of artificial intelligence (AI) such as chatbots could make insurance agents superfluous - or at least relieve the insurance industry.

"Artificial intelligence" has developed into a buzzword in numerous industries, as it can change the working and living environment permanently - including where large amounts of data have to be recorded, structured, analyzed and utilized.

However, this is often more difficult than expected, for example in the practical implementation of AI in autonomous driving, which, despite enormous investments, is still far from being suitable for everyday use.

More sales per customer possible through the use of AI

Large amounts of data are also generated in the finance and insurance business.

Here, AI holds unimagined potential.

For example, it could help to make more effective use of thousands and thousands of rarely used contacts with insurance customers in order to generate more sales or to win new customers.

Insurers and banks seem to have recognized the potential by now and want to invest more in the digitization and automation of their processes.

In fact, only a minority of companies currently use AI applications.

Only 5 percent of banks and 6 percent of insurers came into contact with their customers via AI-supported applications, as a study by Capgemini published at the end of 2020 shows.

A finding that the analysts at Capgemini attribute primarily to resistance among the company's management staff.

Many managers are still slowing down the process

And where financial service providers increasingly used AI-supported applications under the pressure of the corona pandemic in the past year, they apparently did not meet the expectations of customers: The study comes to the conclusion, among other things, that a good half of customers do not get by with added value AI-supported digital contacts with your bank or your insurer.

One reason is, among other things, that they lack the "human-like experience" with AI-supported chatbots.

So better back to "Mr. Kaiser"?

Probably not.

The potential benefits of mature AI applications are too tempting for banks and insurers.

According to the study, financial service providers who use AI in customer contact have reduced their operating costs by 13 percent and increased sales per customer by an average of 10 percent.

These are not outstanding key figures, but they could increase with greater customer acceptance.

In any case, consumers seem to be open to the new technologies: 78 percent of the customers surveyed are willing to use more contactless interactions via voice assistants, facial recognition or apps - and this does not mean contactless payment via internet banking or cards.

Where insurers can score with AI

If insurers are among the protagonists as particularly suitable for AI applications: Where specifically can this technology then be used, where do experts see potential?

Here are a few examples:

Chatbots

A number of insurers already use these text-based dialogue systems in customer contact to relieve employees and accelerate processes.

The chatbot is controlled by voice input, but mostly via the keyboard, and takes on less complex support tasks.

The bot, so the claim, should be able to process several inquiries at the same time.

On the one hand, chatbots are based on linguistic keyword analyzes.

On the other hand, the bot must be technically "fit".

The system must, so to speak, "know" how to deal with a claim report and which parameters it has to query for processing from the customer, explains Jan Langkau from Adesso Insurance Solutions in Dortmund.

Chatbots are badly received by customers if they don't understand the request.

Then the system should be handed over to a real employee - if he or she is currently on duty.

Automated damage assessment - and the tricks

AI applications can help to quickly determine the damage in an accident and also to process it.

A simple cell phone photo, taken at the scene of the accident and transmitted to the insurer, should be enough.

Image analysis and image forensics technologies check the authenticity of the image material.

Was it possible that the photo was edited and the damage to be regulated was made larger than it is?

AI should help identify this and thus prevent attempted fraud.

This modern technology should also enable the amount of simple property damage to be determined quickly.

Self-learning software, which is fed with millions and millions of pictures of damaged cars, helps to quantify the repair effort in a short time.

In the best case, the system shows the way to the nearest authorized workshop or even books an appointment there.

Allianz customers, for example, would be informed of the next steps within an hour via the "Allianz Schaden Express" app based on mobile phone photos submitted or, thanks to the AI ​​application, they would also receive a direct repair approval.

Telematics tariffs - if you drive carefully, you pay less

In motor vehicle insurance, applications are already in use that monitor the customer's driving behavior in detail and in this way create an individual risk profile that then has a decisive influence on the insurance premium.

Such telematics tariffs ("pay-how-you-drive") are offered by Allianz, their fiercest competitor Huk Coburg and, more recently, Ergo Versicherung.

You lure the customer with bonuses and premium savings.

Coupled with an app, the customer receives feedback on his driving behavior in real time and can benefit from other services from the provider.

Digital insurance advisor

So-called robo advisers were first used in the financial sector.

In the meantime, some insurance companies also use such digital consultants, which are usually developed by fintechs for the insurance industry and banks.

In the background, computing algorithms check the risk status and the degree of coverage of a potential customer.

The digital advisor compares, according to the promise, products available on the market and also suggests alternative policies - thanks to AI, everything complies with the law and according to the company's advisory rules, insure the advocates of such applications.

Ideally, the customer receives the suggestions on a mobile device and can then complete them if he so wishes.

The service for the customer will be significantly improved, says Langkau from Adesso Insurance.

While the digital helper takes on simple product recommendations, the real flesh-and-blood consultant can concentrate on customers and products, for whose conclusion more advice and personal contact are crucial.

AI and dark processing

"Dark processing" in insurance?

That sounds almost conspiratorial, but it is anything but mysterious.

A competitive insurer strives to automate its processes as much as possible, to "process them in the dark", as it is said in insider speech.

But despite all the portals and apps, data and documents are still unstructured, incomplete and in very different forms (fax, email, handwritten letter) at the insurer and cost the clerk valuable time (scanning and digitization, among other things).

Insurers help themselves here with systems of optical text recognition.

But only an AI that records and understands the content of a document in a self-learning manner can structure the mass of data and then make it available for further automated processing.

That reads easily, but in this specific case it is anything but banal.

A dark processing of all relevant insurance documents would then be the Kaiser discipline.

rei

Source: spiegel

All news articles on 2021-01-28

You may like

Tech/Game 2024-03-09T01:47:24.568Z

Trends 24h

News/Politics 2024-03-28T06:04:53.137Z

Latest

© Communities 2019 - Privacy

The information on this site is from external sources that are not under our control.
The inclusion of any links does not necessarily imply a recommendation or endorse the views expressed within them.