# E-book pricing & bundles for niche non-fiction sold direct (2026)
Harvested 26 Oct 2026 (skills harvest).

## Practitioner testimony and data
- Seeley James (ALLi, price tests in $1 steps from $2.99 to $8.99, boxed set $4.99–$9.99): no significant sustained difference in units/week between $2.99 and $6.99; a sustained 28% volume drop at $7.99 not made up in royalties; the same royalty at $6.99 as at $5.99; a three-book boxed set "doing well at $9.99". Every price rise caused a 30–60% day-one unit drop that recovered in royalty by day 5 and in units in about two weeks -> never judge a price change in its first two weeks. Every price DROP caused a small buying surge (wish-listers notified). (selfpublishingadvice.org)
- David Gaughran: price and value are different; "a lot easier to sell a book at $2.99 than $9.99"; Smashwords surveys repeatedly put $2.99–$4.99 as income-maximising for retail fiction; he has seen little data for higher prices. Note: that is store fiction; specialised non-fiction readers pay for specific information. (davidgaughran.com)
- Direct-sale economics: Payhip free plan takes 5% per sale (payhip.com). Designrr's "$47 direct vs $9.99 Amazon" example is a vendor illustration, not a case study (not checked).
- Tale-forge guide puts non-fiction at $7.99–$14.99 (guide, not data; not checked).

## How-to
1. Anchor with a bundle: list the bundle first ("all N reports"), priced at about 2–2.5x a single title, so singles read as the low-commitment option. Seeley James's box set held at $9.99.
2. Keep singles in impulse range (about £3–£7 for a 30–80 page heritage report) and test one step at a time; hold each price for 4+ weeks; ignore the first 14 days after any change.
3. A free entry book (or free first chapter) is the list-builder; the paid books are the second step. "A read is a conversion, a sale a stronger one" (the project lead's funnel).
4. Use price drops as events: a time-boxed reduction announced to the list produces the surge Seeley James saw; one per quarter at most, never to someone who already bought.
5. Cross-sell on the thank-you page/receipt: "readers of X also read Y" with the bundle upgrade (pay the difference) where the checkout supports it.
6. Measure per-visitor revenue on /ebooks/ (sales £ / unique visitors), not units alone.

## How we apply it to Cedwyn
- Baseline is 0 sales: first fix is not price but traffic and the free-read step. Do not change prices before ~200 visits to /ebooks/ have been logged at the current price.
- Proposal for the project lead (PROPOSED, not applied): one "Complete Great House Farm archive" bundle shown first on /ebooks/, priced ~2–2.5x a single; a free short e-book (e.g. the 1824 plan 19 story) behind the email signup; singles kept in the impulse band.
- Check current prices on /ebooks/, Stripe and Gumroad (the project workspace) before writing any pricing copy; never quote a price in an email that the shop doesn't show.
- Log every price change with date in the project workspace; compare 4-week windows excluding the first 14 days.
- Weekly promo to subscribers (max once a week) can use a time-boxed price drop on one title; exclude buyers via the purchases list.

## Sources
- selfpublishingadvice.org
- davidgaughran.com
- payhip.com
- designrr.io (vendor; not checked)
- tale-forge.com (guide; not checked)