Cold Email

Cold Email

A cold email is an email sent to a stranger who has had no prior contact with the sender and has not requested anything. It is the classic tool in sales, is subject to strict legal rules, and is nowadays often written by AI software.

A cold email is an email sent to someone who doesn’t know the sender. The recipient hasn’t ordered anything, hasn’t requested anything, and hasn’t given any consent. Nevertheless, the message lands in their inbox, usually with a sales offer or a request for a conversation. The English word “cold” means here: the relationship is still cold, there is no shared history. The opposite would be a message to someone who reached out themselves or subscribed to a newsletter. Cold emails are not spam in the narrow sense, because they are sent individually to selected people. The line between the two is, however, thin.

Why companies write to strangers' inboxes

For many companies, the cold email is the cheapest way to reach new customers. An ad costs money per click, an email costs almost nothing. Especially in business-to-business dealings, the practice is standard. A software provider deliberately writes to the purchasing managers of two hundred mid-sized companies. A few of them reply, and from that, contracts spanning years emerge.

The typical numbers explain why so many emails are sent. Open rates of 30 to 50 percent are considered decent, while reply rates sit at one to five percent. Anyone wanting ten conversations therefore quickly needs several hundred contacts. This exact calculation has given rise to an entire industry. Providers sell address lists, sending software, and templates for the perfect subject line.

The downside is borne by the recipient. If every company runs the same calculation, inboxes overflow. That’s why there are tight legal limits in Germany. Advertising emails to private individuals without their consent are generally impermissible, and even with companies there needs to be a plausible interest in the offer. Violations can result in formal warnings and end up being expensive.

From finding an address to the follow-up email

It starts with research. The sender looks for people for whom their offer might be a good fit, for example via career networks or company directories. Role and context matter: who decides on the purchase, and what is currently occupying this company? From this information, a list emerges, referred to in industry jargon as leads, meaning contacts with sales potential.

The email itself follows a fixed structure. A short subject line, an opening sentence with a personal reference, then the problem one wants to solve. At the end there is a single, clear request, usually a short phone call. Long texts don’t work well; five to eight sentences is common. Anyone who doesn’t reply receives a reminder after a few days, the so-called follow-up email.

This is where AI comes into play. Language models can automatically formulate a fitting opening sentence from a target person’s profile. The same basic template is thus varied individually a thousandfold. This increases the reply rate, but also lowers the barrier to mass sending. A common misconception is that personalized phrasing alone already makes it legally permissible. What remains decisive is consent, not the tone.

Recognizing cold emails in your own inbox

Anyone who posts their own address publicly online will sooner or later receive such messages. Typical features are the first name in the subject line, a reference to a supposedly shared topic, and a scheduling link at the end. Internship and job inquiries from students to companies are also cold emails, just without a sales intent. The principle is the same: a stranger writes and hopes for a reply.

In business news, the term usually comes up in connection with sales software. Start-ups offering automated outreach regularly raise large amounts of investment. At the same time, email providers such as Google and Microsoft are tightening their filters against mass sending. Together, both factors turn cold emails into an ongoing race between senders and protective mechanisms.

Subscribe free. Unsubscribe the second it sucks.

High-signal news across AI, business, UX, and tech. Every morning.