If your reply rate is near zero, the copy is probably not the problem. Writing well to people with no reason to care produces well-written failure. Almost everything that determines whether outbound works is decided before a word is written, in who ends up on the list.
"Marine businesses in New Zealand" is a segment. It tells you where to look and nothing about who to write to. "Independent boatbuilders in Northland with fewer than fifteen staff, no online booking, and a backlog long enough that they turn work away" is a profile. It tells you what to say, what to offer, and — crucially — who to leave out.
The difference shows up immediately in the writing. Against a segment you can only write generalities. Against a profile you can write the specific sentence that makes someone reply.
From customers you already have, if you have any. Look at who bought fastest, who was easiest to serve, who paid without friction, and who is still around. Then find what those have in common that is observable from outside — because a criterion you cannot check on a company's website cannot be used to build a list.
If you have no customers yet, the profile is a hypothesis. Write it down anyway, send to a small batch, and revise it based on who replies. The people who answer are telling you what your profile actually is, which is usually not what you assumed.
Take your profile and apply it to twenty real companies. If you cannot confidently include or exclude each one in under a minute using only their public information, the profile is too vague to build a list from — and too vague to write a specific email from.
Narrow enough that you could write a genuinely different email to this group than to any other, and broad enough that enough companies exist to build a list. If your profile matches thirty companies in your market, it is a niche worth serving personally rather than a segment worth automating. If it matches thirty thousand, it is not a profile.
Yes, and running two or three separate campaigns against separate profiles works well. What does not work is merging them into one campaign with one message — the message ends up written for an average buyer who does not exist.
The replies tell you. If the people answering share a characteristic your profile does not mention, add it. If a group you targeted heavily never answers, remove them. Treat the first hundred sends as research, because that is what they are.
Fewer prospects, usually more revenue. A hundred well-matched companies with a real trigger outperform a thousand vaguely relevant ones, and they cost less to reach — in sending volume, in domain risk, and in the hours spent writing to people who were never going to answer.
The Cold Outreach Machine finds the companies you sell to, writes a different email for each one, and sends them on its own — paced, inside working hours. US$299.99, paid once, live setup call included.