How does AI-powered inbound mail automation work?

Dirk Schallhorn
Executive Key Account Manager Insurance, BSI

Good old e-mail. It has been declared dead often, yet it continues to be widely appreciated and used intensely. Whether it is marketing, sales or service – it cannot be done without e-mail communication. Even new channels such as chatbots, self-service portals, WhatsApp & the like have hardly changed that to date. At the AMC (Assekuranz Marketing Circle) focus day called “E-mail Marketing and Automation,” experts discussed the latest developments in e-mail marketing. We at BSI were there, too, with an impromptu presentation on the topic of inbound mail automation.

E-mail marketing continues to be an attractive way to reach prospects and customers quickly, easily and cost-effectively. However, spam, phishing and information overload, plus data protection regulations, present major challenges in this marketing area. What it comes down to additionally: perfect orchestration, the right timing, (hyper-)personalized content, the right response to each customer trigger, automatic processes, and the list goes on.

Did you know?
Only 18% of companies rely on e-mail automation campaigns. 27% use message personalization as a tactic in e-mail marketing. (Hubspot, 2021)


No reply is not an option

In a keynote speech, Dominique Maron, customer experience expert, and the data scientist Patrick Frank shed light on the topic from a slightly different perspective: inbound mail automation using Artificial Intelligence (AI). They discussed what added value automation and AI applications can provide to incoming customer e-mails. The speakers emphasized that no-reply addresses are not an option and indicated that they would hope companies want to interact with their customers and process customer inquiries in a way that makes their customers happy, lighten their employees’ workload and accelerate their processes. Cost savings are also included.

A customer example

Using the imaginary customer Lena Meier as an example, Patrick Frank and Dominique Maron showed convincingly what kind of data can be extracted from a single e-mail in an automated manner.

Lena Meier sends an e-mail to customer service and addresses several concerns: She has moved and includes her new address. At the same time, Lena also asks whether her insurance company still sponsors the Reitverein Köln (Cologne Riding Club). She refers to the company’s earlier ticket campaign and asks for two tickets for her daughters. She also speaks highly of the earlier event.

What happens next?

With the help of state-of-the-art machine learning models, the CRM extracts implied information such as mood or language style. In addition, it recommends the next best actions based on recognized content and interests or creates business cases. The information generated from the customer e-mail is available to the sales rep or customer service rep in real time in the 360° view of the CRM.

Specifically, this means:

  • The customer inquiry is automatically summarized;
  • Lena Meier’s e-mail is saved in the CRM;
  • AI automatically detects the mood of the e-mail;
  • The address change process is initiated;
  • The next best action, “Event Invitation,” triggers a suitable customer journey;
  • “Sport” is set as an interest of Lena Meier’s, which is valuable information for future interactions with her;
  • Appropriate personalized response suggestions are generated to reduce the sales rep’s or customer service rep’s workload;
  • The sales rep or customer service rep reviews and edits the reply e-mail if needed; the e-mail is sent and recorded in the CRM.

So, Lena Meier receives a personalized e-mail from her sales rep within a noticeably short time confirming the address change and the two tickets for the desired event. In addition, she can look forward to more customized offers in the future.

There are no limits to your creativity

Yet AI can do even more than that. The applications are almost endless, from customer value determination and lead scoring to fraud detection. We are always interested in knowledge sharing and are constantly looking for new use cases. What kind of cases are you most interested in? Please feel free to contact us.


Would you like to learn more? We look forward to talking with you.

Send inquiry now

Next Best Action: BSI

The next best action: Network your goals with BSI.

Oliver Hechler, Creator of Customer Delight

+49 89 189 170 912

BSI Customer Suite

Provide a truly compelling customer experience with seven products – all in one solution.