How Many Drafts Does a Novel Actually Take?

When I first heard people say a novel usually takes five to seven drafts, I assumed they were speaking figuratively.

Seven drafts sounded excessive. How much could you possibly revise the same story without just… rewriting it entirely?

As it turns out, that assumption was doing a lot of quiet work.

I recently counted the number of drafts Whispers in the Dark went through on its way to the current manuscript. I had to count because I didn’t label them “Draft 1,” “Draft 2,” and so on.

I used version numbers instead.

Semantic Versioning, But for a Novel

This is probably unsurprising given my background, but I didn’t think in terms of drafts so much as iterations.

Minor changes — tightening prose, adjusting pacing, clarifying motivations — bumped the version by a decimal point:

  • 1.1
  • 1.2
  • 1.5

Major structural changes — reordered arcs, replaced core ideas, killed entire sections — got a new leading number:

  • 2.0
  • 3.0

By the time I reached what I now consider the “real” version of the book, the sequence looked something like this:

  • 1.1 → 1.2 → 1.5
  • 2.1
  • 2.6 → 2.7
  • 3.0 → 3.1

Depending on how you count, that’s eight or nine drafts.

Which puts me squarely in the range I originally thought sounded unrealistic.

The Realisation I Missed Early On

What I hadn’t understood at the start was that drafts aren’t revisions of the same thing.

They’re different solutions to the same problem.

Version 1.x wasn’t “an early version of the final book.” It was an exploration of what the book might be.

Version 2.x wasn’t polish. It was a recognition that some earlier assumptions didn’t hold up, and the structure needed to change to support what the story was actually trying to do.

By the time I hit 3.0, I wasn’t rewriting for the sake of rewriting. I was converging. The big questions had been answered. The remaining work was refinement, not reinvention.

Seen that way, the number stops feeling large.

Iteration Is the Point

I think part of why the “seven drafts” advice sounds intimidating is that it’s often framed as repetition: doing the same thing over and over until it’s perfect.

That wasn’t my experience.

Each major version existed because the previous one taught me something concrete:

  • a constraint I couldn’t ignore
  • an idea that didn’t scale
  • a character arc that broke under scrutiny
  • a concept that worked emotionally but not structurally

Once I saw the issue clearly, the next version wasn’t incremental — it was decisively different.

Iteration wasn’t a sign I’d failed to get it right the first time. It was how I figured out what “right” actually meant.

Faster Next Time? Probably. Fewer Drafts? Maybe Not.

I’m hopeful that the next project will converge faster. I recognise the phases now. I know what an exploratory draft feels like versus a structural one. I’m less likely to mistake momentum for correctness.

But I’m not convinced that means fewer drafts.

It might just mean the same amount of learning, bundled more efficiently — bigger jumps, fewer decimals.

Going from 1.0 to 2.0 faster doesn’t mean skipping the work. It just means recognising sooner when a version has given you everything it can.

An Open Question

I’m curious how others experience this.

Do they find that they tend to land in roughly the same number of drafts each time, regardless of experience? Or as they improve, do some of those revisions collapse together — fewer versions, but bigger jumps between them?

And if you think in terms of drafts rather than versions: at what point does a revision stop being “the same book” and start being something fundamentally new?

I don’t think there’s a correct number. But I do think counting — or at least naming — the iterations can make the process feel a lot less mysterious.

Enjoyed this? Get the science behind the story → Join the list