From targeting to iteration; how we use cold emails to get paying clients

I run a lead generation agency and serve multiple clients from all around the world. We get around 3–6 new closed clients on average and a reply rate between 3% and 20%. I tell you this because I want to clarify that cold emails work when you treat them right. Every channel works when you treat them right.

It's a vehicle to get clients, not the ultimate solution for everything. But doing it right will be the most cost efficient decision for your business. Here is how we operate with our clients to get replies:

1. Target over everything
   80% is targeting, 20% is messaging. The best message, the right words, and even personalization don’t make sense when the targeting is not right. Apollo or other databases are not consistent in their data, even the best keywords and parameters will leave you with 20–50% mismatches that you have to clean up before writing any messages.

2. Quality over quantity
   Instantly and other sending tools make it look very easy to blast out thousands of emails per day, but the problem lies in the quality of the outreach, the research and reason for a relevant message to the right audience. Therefore, we built our own client specific qualification system that researches the company and person, gathers more information about their current situation, and closes the gap between what my clients offer and what they might need. For web design services, we search for the current website performance data, for B2B SaaS, we look for current hires and technologies used. It's client specific but highly relevant.

3. Personalization with care
   Nobody wants to read a "personalized" email (–icebreaker). They don’t want to read cold emails at all. That’s a difference nobody really addresses. They don't know or owe you anything. In fact, they don't care and are potentially annoyed about the unsolicited email. So your goal is to give them a reason why they should care. You should provide them with information they might be missing and present a solution for that hint.

4. Iteration and testing
   A static system is blind and full of failures. Only relying on thinking will not be enough these days. So, in addition to the automated qualification and personalization system, we always implement a data analyst, an automated process that collects performance data and fully automatically writes daily reports in Word format that I can read in the morning for every client. These include copy suggestions, campaign performance, and actionable steps to take. This way, the system learns by itself, and so do we.

Don't get me wrong, there are many ways to Rome, but from experience, these four steps lead to successful campaigns. This text box is not enough to go from vague to specific, but I'm happy to discuss things further.