Writing Time Estimator
"How long does it take to write a blog post?" has no single answer, but it does have a structure. Break a piece into stages, put your own hours against each one, and the total becomes something you can plan and price. This estimator does that for a piece written with and without AI tools.
Why the defaults look the way they do
- Drafting is where AI saves the most. A model can produce a workable first pass quickly — but it still needs restructuring and rewriting around the angle and the writer's own material.
- Fact-checking goes up, not down. AI drafts state errors fluently and sometimes invent sources. Every claim needs checking against a primary source, which takes longer than checking your own notes.
- Briefs and decisions don't get faster. Agreeing who the piece is for and what it must achieve is a conversation, not a generation task.
- Research gets a modest boost. Summaries help you triage long sources, but you still have to read the ones you cite.
The result is usually a meaningful but not dramatic saving. Claims of several-fold productivity gains generally measure drafting speed alone, or accept a lower quality bar.
How to measure your own numbers
- Pick three or four upcoming pieces of a typical type.
- Log time per stage with a simple timer — including interruptions for client questions and revisions.
- Divide finished words by total hours: that is your real throughput, and the figure to use in the rate calculator.
- Repeat after changing your process (for example, introducing AI for research triage) to see what actually changed.
Using the estimate in a quote
Quote per piece rather than per hour where you can, using the estimate to set the price and a clear scope to protect it: number of revision rounds, who supplies sources or interview access, and what happens if the brief changes. If a client asks for AI-assisted work to be cheaper, show them the stage breakdown — the saved hours are real but smaller than they expect, and the review work is what keeps their content accurate.