5 min read
It is 9 p.m. on a Tuesday, and Dana is still at her desk with fourteen browser tabs open, trying to understand a client’s industry well enough to write about it convincingly by morning. The writing itself will take two hours. The getting-ready-to-write has already eaten four. If you have ever freelanced, you know this is the part nobody warns you about: the work behind the work.
Dana is an illustrative composite, not a specific person. The details below reflect how a lot of solo copywriters are actually reworking their days with AI, and the numbers are realistic examples rather than audited figures.
The writer, and the trap she was in
Dana runs a one-person copywriting business. She writes website pages, email sequences, and the occasional launch campaign for small companies, and she is good at it. Her problem was never the writing. It was that every new client came with a mountain of unfamiliar context, a new industry, a new competitor set, a new pile of the client’s own past material, and she had to climb all of it before she could write a single confident sentence. Booked solid on paper, she was quietly capped, because the research tax meant she could only take on so many new clients before her evenings disappeared.
What was actually costing her
The stakes were not dramatic, which is exactly why they were easy to ignore. No single project was a disaster. But the pattern added up to real money. Dana estimated she spent roughly half of every project on research and orientation before writing began. That is half her billable capacity going to work she could not really charge for at full rate. It also meant she turned down projects in unfamiliar fields, because the ramp-up felt too steep, and quietly watched a competitor take them instead.
The fixes that did not work
She tried the obvious things first. She built templates, which helped with structure but did nothing for the research. She tried skimming less, and produced weaker work. She tried charging a research fee, and lost two prospects to sticker shock. None of it touched the real bottleneck, which was the sheer time it took to turn a stranger’s industry into something she understood well enough to write about with authority.
The setup she landed on
What finally moved the needle was not one tool but a small, deliberate sequence. Dana split her work into two jobs that used to blur together: finding out, and writing.
For finding out, she started every project with a research assistant like Perplexity, asking it to summarize the client’s industry, name the main competitors, and pull out the language those competitors use, each answer arriving with sources she could click to verify. What used to be four hours of tab-juggling became about forty minutes of guided reading, because she was checking a drafted summary instead of building one from scratch.
For the writing, she kept it firmly in her own hands, but used a writing assistant like Claude or ChatGPT as a sparring partner: feeding it the research and her rough outline, asking it to poke holes, suggest angles, and draft throwaway versions she could react to. She never shipped its words. She used it to get unstuck faster. If you are weighing which of those assistants to pay for, our comparison of ChatGPT, Claude, and Gemini covers the trade-offs.
The last piece was reach. Once a project was done, Dana used the same approach behind our guide on turning one blog post into a week of content to spin each finished project into a short case note and a few social posts, so her own marketing stopped being the thing she never got to.
What changed
In Dana’s case, cutting the research phase roughly in half did not just save hours. It changed what she could say yes to. She began taking projects in industries she would have refused a year earlier, because the ramp-up no longer scared her. Her per-project profit went up, not because she raised rates, but because far less of each project was unpaid preparation. And the evenings came back, which was the part she actually cared about. None of this required her to become technical. It required her to stop treating research and writing as one undifferentiated slog and hand the first half to a tool built for it.
What another solo owner can take from this
Separate finding-out from making. Most creative solo work hides a research phase inside it, and that phase is usually the one an AI tool can compress the most. Name it, time it, and hand it over first. The making, the part clients actually pay you for, stays yours.
Use AI to react, not to replace. Dana’s output stayed hers because she used the assistant to get unstuck, not to write the final words. That single rule is what kept her voice intact and her clients happy. The same discipline let a solo web designer cut his project overhead from six hours to two without cheapening the work.
Reinvest the freed time on purpose. Time you win back leaks away unless you point it somewhere. Dana pointed hers at her own marketing and at harder, better-paid projects. Decide in advance where yours goes, or it will quietly refill with busywork.
Try the two-list test this week
Take your next project and, before you start, write two lists: the finding-out tasks and the making tasks. Hand the first list to a research assistant and keep the second for yourself. See how much of the finding-out you can turn into checking-a-draft instead of building-from-nothing. If it saves you even an hour, you have found the loose thread that unravels the whole time trap. Which half of your work has been quietly eating your evenings? Start there, and hand it over first.
