The five-year math
Take a typical mid-tier writing subscription at $15/month, which is roughly the going rate for the major cloud-based author tools in 2026. That's $180/year. Over a working author's typical project arc, first draft, revisions, publication, second book in the same world, you're looking at three to five years of paying for the tool.
- Year 1: $180
- Year 3: $540 cumulative
- Year 5: $900 cumulative
- Year 5, with a 30% subscription price increase mid-way through (which is normal): about $1,100
For comparison, Edda is $39.99 for Windows and Linux, $39.99 for macOS, $39.99 for Android, or $79.99 for the complete bundle. One time. Across the same five-year horizon, Edda costs you 5% of the subscription tool. That's not a marketing claim; it's arithmetic.
The cost the math doesn't capture
The bigger problem isn't the dollar number, it's what the subscription model does to your incentives:
- You can't stop. Cancel the subscription and access to your documents downgrades or disappears. A one-time purchase lets you put a project on the shelf for two years and pick it up exactly where you left off.
- Pricing power is one-way. Once your manuscript is in their format, the vendor's pricing decisions are yours to absorb. You have no walk-away leverage.
- Feature lock isn't the lock. Document lock is the lock. Many subscription writing apps will let you export your manuscript on cancellation, but the worldbuilding system, the timeline, the character relationship graph, all the structured data you built for years, those usually don't export cleanly.
When subscriptions do make sense
To be fair: subscriptions aren't inherently wrong. A subscription pays for a service that genuinely costs the vendor money to run continuously, cloud storage at scale, real-time collaboration, ongoing server-side compute. If you actively use those server-side features every day, the math can work out.
The issue is that most authors don't use those features. They use a text editor with some structure on top. Paying a monthly fee for that is paying for a feature you don't need.
A test for whether you should switch
If you can answer yes to all of these, a one-time-purchase tool is almost certainly the better deal:
- You write on one or two machines, not eight.
- You don't need real-time collaboration with a co-author.
- You can imagine taking six months off and coming back to the same project.
- You want every helper inside the editor, spell check, style flags, prompts, readability, to run on your own machine.
That's most working authors. The subscription model survives largely on default inertia; running the math is usually all it takes to change a workflow.