It was Tuesday morning. My inbox was, as usual, a jumble of bills, advertising offers, and a couple of “urgent” letters from colleagues. One was from a service I had recently subscribed to. I opened it. Everything was neat: the greeting, the structure, even the emoji at the end. But… it was empty. Not a single word to hook onto, to recognize myself in. It was as if the letter had been written for thousands of people like me, with my name simply inserted.
In marketing, the word “scalability” is perceived in a hypnotic way; it predetermines potential and inevitable success. This is accompanied by automation, scripts, APIs to LLM — all of which look great on slides. Companies turn to N-iX for generative AI consulting in order to achieve a “universal brand voice” across all channels. And yes, the technology there is impressive, but there is a catch — standardized texts cease to be personal.
The fact is that customers of expensive products do not come for the speed of response. They do not pay for milliseconds of email delivery, but every customer is ready to open their wallet the moment they realize, “I have been noticed” or “I matter.” It’s like the difference between “Thank you for your order” and “Mark, I hope you’ve already opened the coffee you ordered last night.”
How Not to Ruin Everything in 0.3 Seconds
Sometimes you open an email and realize after just two lines that you don’t want to read it. It takes less than a second to understand who the email is addressed to (studies show that it takes about 0.3 seconds ). During this time, the potential customer’s brain decides: “This is about me” or “This is about thousands of people like me.”
Even if there seems to be personalization: a name in the header, a product from the last purchase. But AI generation can give away the tone of the letter if it is sterile, like in the instructions for a vacuum cleaner. The customer feels that it is not about them.
And in niches where trust is currency (B2B SaaS, premium services, consulting), such a “sterile” voice = no deal. Nielsen Norman Group noticed that if a letter deviates from the usual tone, the response rate drops by 14–22%. It doesn’t matter what the trigger is: excessive politeness, dry generalizations, or strange wording. People don’t analyze, they just intuitively understand that they are being spoken to using a template.
Where AI Is Appropriate, and Where It Is Poison
You may object: “People are getting used to AI content.” They are getting used to it, but not everywhere. Yes, in tech updates or delivery notifications, it doesn’t matter who wrote it. But in a letter where a $200k contract is being decided, a template response is not just a mistake. It’s a signal: “We didn’t bother to talk to you personally.”
Experience shows that customers forgive a lot : typos, minor exaggerations, even awkward jokes. But they don’t forgive the feeling that they’ve been funneled into a CRM funnel and run through a script. And that’s when a brand loses not only the deal, but also respect.
And this is where the difficult conversation begins.
Stories That Lost Their Voice
Case 1: Restaurant chain in Boston
They implemented AI to process reservations and post-sale mailings. The first two months were a delight: everything was faster, letters were sent on time, and the style was consistent. In the third month, repeat reservations began to decline. A survey showed that 27% of regulars stopped coming because it had become “soulless.” One wrote: “I celebrated my birthday here for six years, and this year I received a dry thank you. There wasn’t even a name.”
Case 2: SaaS platform for data analysts
The AI chatbot responded to customers according to a script, without taking into account the history of communication. As a result, users began to write directly to managers on LinkedIn, bypassing support. The NPS metric fell by 19 points. The CTO admitted: “We saved 40 hours a month, but lost the loyalty of key accounts.”
What’s the mistake?
AI has become your replacement, not a tool. But it cannot be allowed into areas where engagement is more important than speed and accuracy.
On a technical level, it often looks like this: the bot is simply integrated into the CRM, and it responds to everyone equally quickly, but without taking into account that someone’s subscription is about to expire, while someone else has simply forgotten their password. This is not specified in the SLA, and priorities become blurred.
Why Is This Important For IT Directors And Product Managers?
When communicating with customers, you are responsible for more than just features and bug fixes. You are responsible for the tone of the brand at those points where the customer decides whether to stay or leave.
Alternative:
- Hybrid model: AI generates a draft → a person adds context (real details, references to past interactions).
- A clear map of areas: where AI is acceptable and where human involvement is required.
Experts in customer experience, digital communications, and generative AI training at N-iX talk about the loss of authenticity in customer communication due to excessive automation and the use of generative AI as a strategic loss. This happens when you’ve done everything right in terms of efficiency, but ended up closing the door on those who have been with you for years.
Conclusion
Imagine that your future success or failure depends on the answer to one simple question: After reading your letter, what will the customer remember — the speed of your mailing or the fact that you were the only one to call them by name when they made a repeat purchase?
Check your inbox. And be honest: was this email written by a human or a bot?
What if the customer doesn’t care? They might. Until they encounter your competitor who writes as if they know them personally. To know where to grow next, check your AI touchpoints with generative AI consulting; this will help you find out where your voice is heard and where it is just a faceless template.
The rest is up to you.
