UPSET CUSTOM- ERS WRITE LONG EMAILS; FIGURING OUT WHAT THEY’ RE ASKING FOR CAN BE LIKE TRYING TO FIND THE PROVER- BIAL NEEDLE IN THE HAYSTACK.
2. CUSTOMIZING RESPONSES BASED ON EMAIL TEMPLATES
To maintain efficiency and consistency, contact center agents use templates or macros to respond to customers. These prewritten“ form letters” can be helpful aids, but agents must customize the templates to create worthwhile responses, ones that customers can trust.
Some templates include prompts for customizing, such as:
•“ We’ re sorry to learn you’ re disappointed with the [ insert product name ] that you purchased on [ date ]…”
•“ Thanks for contacting us. [ Insert empathy statement.]”
But customizing something that’ s already written can be a lot harder than it looks. While it’ s relatively easy for agents to drop in a purchase date or a product name, weaving an expression of authentic empathy into a template or adding a correction of a customer’ s misunderstanding requires substantial editing.
Agents must then be willing and able to add freetexted sentences into an already-written template. It’ s hard work that has to be done quickly and separately.
GenAI can make it easier to prepare a customized response based on an email template. The agent would include the customer’ s incoming email and the email template in the prompt. Yes, the prompt would be quite long, but most of the content would be copy-and-pasted.
Here’ s how to prompt the AI:
“ Using this email template [ insert the email template here ], write a response to this customer’ s email [ insert the customer’ s email here ].
Customize the email template by using details the customer has shared in their incoming email. Base your response on the email template, but do not use the template as-is.
Delete any information from the template that’ s not relevant to the customer’ s question and add customized information to make the response particular to the customer.
Use the customer’ s details about what they purchased, what they want from our company, how they’ re feeling, the questions or complaints about their purchase, or the type of help they’ re looking for.”
3. CHECKING OR CHANGING THE TONES OF RESPONSES AI tools can help agents improve more than what they’ re writing to customers; the tools can help agents adjust how they’ re writing.
Shaping the tone of an email response can be a fiddly, time-consuming task, but GenAI can help with prompts like these:
" Revise this email response to give it a friendlier, more personal tone.
“ Check the tone of this email response to ensure it is understanding without being conciliatory because we cannot agree to the customer’ s request for a refund.
“ Check the tone of this email to ensure it complies with our company’ s Brand Voice and Tone Guidelines [ link to Guidelines document ] and edit the tone if it does not comply. List the changes you’ ve made to the email.”
4. AS A READING TOOL
Reading hundreds of incoming emails is difficult. Because customers are often asking the same questions or making the same complaints, their emails blend together.
The more emotional the customer, the more they write. Upset customers write long emails; figuring out what they’ re asking for can be like trying to find the proverbial needle in the haystack.
AI tools can digest detailed information quickly and prepare the agent to respond accurately and with less effort. Here’ s how:
• Summarize a long email from a customer by listing the customer’ s questions. Some customers write 10 paragraphs when one would be enough.
Before replying to one of those 500- word email treatises, an agent could prompt GenAI to list the customer’ s questions, both stated and implied.
WRITING WELL
UPSET CUSTOM- ERS WRITE LONG EMAILS; FIGURING OUT WHAT THEY’ RE ASKING FOR CAN BE LIKE TRYING TO FIND THE PROVER- BIAL NEEDLE IN THE HAYSTACK.
• Summarize an emotional email from a customer by stripping out the feelings and creating a neutral version.
An airline I worked with a few years ago received a 969-word email from a customer. He was heartbroken and furious because problems with his visa and other travel documents meant his international travel was disrupted, and he was temporarily detained at the airport mid-itinerary.
He used emotional phrases like these:--I am extremely disappointed.--I have suffered your abuse of power.--I cannot express how frustrated I was.
--I beg you for moral compensation for all the distress your staff created.
--You have permanently broken my trust.
The heightened emotion in the customer’ s email made it harder for agents to understand what happened and what the airline could do to make things better.
If GenAI had been available, the agent could have told the AI tool to create a summary of the customer’ s email that focused on the facts of the complaint and stripped out the feelings.
Using the GenAI-created summary, the agent could have drafted a response that addressed the problems the airline traveler experienced. Specifically, the problems the airline could solve or compensate for.
MARCH 2026 17