How the calculation works
The calculator starts with the two numbers any Amazon Ads campaign reports every day: ad spend and cost per click. Ad spend divided by cost per click equals total clicks. A $200 budget at $0.40 per click is 500 clicks. That is the size of the audience you bought.
It then applies your click-to-sale conversion rate to get sales. If 10% of clicks become purchases, 500 clicks produce 50 sales. The same campaign at 8% produces 40 sales; at 15% it produces 75. Small differences in conversion create large differences in unit economics because the spend is fixed and the sales scale linearly.
From sales, the calculator derives net profit per sale. This number depends on platform and price point. For Kindle Direct eBooks priced between $2.99 and $9.99, Amazon KDP pays 70% royalty minus a small delivery fee (roughly $0.15 per MB of file size). A $4.99 text-only novel earns about $3.34 net. Below $2.99 or above $9.99, the same file earns 35%, or about $1.74 at $4.99 but $1.05 at $2.99. Add a print cost and the net drops further, because Amazon subtracts print cost before applying the royalty. Kindle Unlimited does not pay per sale at all; it pays per page read from a monthly KDP Select fund, typically $0.004 to $0.005 per KENP page, so a 300-page book read end to end earns roughly $1.20 to $1.50 per borrow. Wide distribution through Apple Books, Kobo, and Google Play pays 70% at most price points without the cliff at $2.99, but the traffic sources and conversion rates differ.
Once net per sale is known, the calculator applies read-through across the series. Each reader who buys book one has your read-through probability of buying book two, then book three, and so on with compounding decay. The total series value per reader is the sum of those expected purchases. A 50% read-through on a three-book series at $3.34 net means each reader is worth $3.34 plus half of $3.34 for book two plus a quarter of $3.34 for book three, or about $5.85 in total. The same 50% on a five-book series is worth about $6.50, and on a ten-book series about $6.67, because each additional book adds a smaller slice. A 60% read-through on the same five books is worth more like $7.50, while a 30% read-through is worth $4.77. The difference between 30% and 60% is not a rounding error. It is the difference between a campaign that loses money and one that funds the next book.
ROI is then computed twice: once for book one only, and once for the full series. Book-one profit is sales times net per sale minus ad spend. Series profit is sales times series value per reader minus ad spend. ROI is profit divided by spend. The gap between those two numbers is why so many profitable series look unprofitable on book one alone and why authors pause campaigns that were actually working.
Platform math: Kindle Direct, Kindle Unlimited, and wide
Kindle Direct is the default for most launch math, but the three platforms pay differently and that changes the break-even cost per click.
Between $2.99 and $9.99, Kindle Direct at 70% pays net equals price times 0.70 minus delivery cost. Below $2.99 the rate collapses to 35% with delivery still deducted. That cliff is why advertising a $0.99 or $1.99 eBook rarely pencils. At $0.99, net is about $0.35. Even a $0.15 CPC with a 10% conversion rate costs $1.50 per sale to acquire, so every sale loses $1.15 before read-through. The only way a sub-$2.99 price advertises profitably is with extraordinary conversion, very cheap clicks, or a deliberate loss-leader strategy where book one is free or $0.99 to feed a higher-priced series behind it.
Above $9.99, Kindle Direct also drops to 35% for eBooks. A $12.99 eBook at 35% nets about $4.54. If your nonfiction eBook is priced there for perceived value, the same CPC that was profitable at $4.99 may still be profitable at $12.99, but the volume at that price point is usually lower. The calculator applies the 35% rule above the threshold automatically so the net number you see for a $12.99 entry is not the $9.09 you would expect from 70%.
Kindle Unlimited pays per KENP page read. The fund value floats monthly and Amazon reports it after the fact, but $0.0043 to $0.0048 has been the typical range. A 250-KENP book therefore pays roughly $1.07 to $1.20 per full read, and a 500-KENP epic pays about twice that. KU borrowers do not generate a sale for conversion rate in the same way; the meaningful rate is borrows per click and completion rate rather than purchases per click. The tool models KU with a $1.00 per-read approximation for quick planning and flags the platform so you know the number is not a royalty. If your KU book is 400 pages and your read-through to a full read is high, the effective net per borrower can be higher than a 35% eBook sale even when the per-page rate looks small.
Wide distribution generally pays 70% without the $2.99 cliff, so a $2.49 wide book is more viable to advertise than the same price on KDP. The tradeoff is traffic: Amazon search clicks convert directly on Amazon where friction is lowest, while Meta or BookBub clicks convert on a landing page with higher drop-off. Keep price and read-through constant and change platform to compare break-even CPC.
If you advertise print, include print cost. A $18.99 paperback with $4.20 print nets $7.19 at 60%, more than twice a $4.99 eBook at $3.34. Enter print cost so you can see which format you want the click to produce.
Read-through by genre and why it varies
Read-through is not a book quality score. It is a genre behavior.
Romance at 50 to 70% from book one to book two is common because series readers are habitual, author brands are strong, and cliffhanger or interconnected-standalone structures create immediate next-book demand. A romance author who sees 45% is not failing; she is inside the band, and the marginal gain from pushing to 60% is worth more than a small improvement in CPC.
Thriller and mystery at 45 to 60% follows a similar logic with character-driven series. A detective or private investigator who returns carries the attachment. Standalone thrillers still generate series read-through when the author brand is the product, but the rate is usually at the low end of the band unless the books share a world.
Fantasy at 40 to 55% reflects long books and slower consumption. World-building investment pays off across volumes, but the time between purchases is longer, and some readers sample rather than commit. Fantasy series with deliberate serialized endings sit at the high end; episodic collections sit lower.
LitRPG at 55 to 75% is an outlier because the structure is serialized by design. A 65% rate on a seven-book series means each ad-acquired reader is worth nearly three sales, which is why high CPCs work there.
Nonfiction at 15 to 30% is low because buyers come for a specific problem. Read-through rises when the books are a deliberate sequence — foundations, advanced, cases — or when a course or newsletter captures the reader outside the retailer.
Literary fiction at 20 to 40% is standalone heavy. Read-through tracks author brand more than series structure. A debut literary novel with a 25% rate to a second book is solid; a 40% rate usually indicates a defined audience like a translated series or a prize follow-on.
Measure your own rate as soon as you have two books. In KDP reports, read-through from book one to book two is units of book two divided by units of book one over the same cohort window, ideally 30 or 60 days after book one sales that originated from ads. Do not use lifetime units of both books without matching cohorts, because backlist sales inflate the numerator with organic readers who never saw the ad. The calculator compounds read-through with decay for longer series — if 60% of book one readers buy book two, only about 45 to 50% of those buyers buy book three — which matches observed drop-off. A flat 60% across ten books is almost always optimistic.
Break-even CPC and the ACOS lens
There are two quick ways to judge a CPC before running full ROI.
Book-one break-even CPC is net per sale times conversion. At $3.34 net and 8% conversion, book-one break-even is $0.267 — pay more and book one loses money. Series break-even CPC is series value times conversion; at $6.50 series value with 50% read-through over five books, the same 8% supports up to $0.52 per click. The gap between $0.267 and $0.52 is what read-through buys.
ACOS is spend divided by attributed revenue; $100 for $300 revenue is 33% ACOS. The console reports book-one ACOS only, but series ACOS — spend divided by series revenue — is the one that pays. Many campaigns run 50 to 80% on book one and 25 to 40% on series. Do not chase low CPC alone: $0.25 at 4% costs $6.25 per sale while $0.40 at 10% costs $4.00. The calculator uses both so you can see why higher CPC can be cheaper.
Common mistakes that destroy the math
Judging ROI on book one sales alone is the most common and most expensive. Authors see a negative book-one profit after $200 of spend and pause the campaign during the learning window. Had they let read-through play out for 30 days, the same spend would have shown a positive series profit. The fix is to judge on series after one full read-through window, not after three days of book-one sales.
Pricing at $0.99 for direct profit is almost never viable. At $0.35 net, even $0.10 CPC with 20% conversion — excellent — costs $0.50 to acquire a $0.35 sale. The loss is structural. If $0.99 is your price, pair it with a higher-priced continuation and treat the negative as acquisition cost.
Ignoring delivery costs at 70% royalty is small but real. A text-only 80k-word novel at $4.99 incurs roughly $0.13 of delivery. A heavily illustrated 250k-word nonfiction file with embedded images can incur $0.60 or more, shaving net from $3.34 toward $2.80. Check your KDP Content Pricing page for the actual delivery cost for the file you uploaded, not the file you edited.
Overestimating conversion is the most optimistic error. A 10 to 15% click-to-sale rate for an exact-match keyword on a strong cover is realistic. Anything above 20% usually means the data set is too small, the keyword is your own title, or the reporting window is too short. Be conservative. It is better to under-project at 8% and be pleasantly surprised than to budget at 18% and stall.
Not tracking read-through separately per book is the silent error for series longer than three. The compounding decay from book two to book three to book four is real. If 60% of book one readers buy book two, expect 45 to 50% of those to buy book three, and 40 to 45% of those to buy book four. The calculator's default decay matches this. If you enter a flat high rate across ten books, you will overstate series value by a full book or more.
Mismatching platform and traffic is the channel error. Advertising a KU title with Amazon Sponsored Product clicks captures Kindle intent efficiently. Advertising a wide title with the same clicks and then sending buyers to a non-Amazon landing page leaks conversion. Keep platform and traffic matched when you compare CPCs.
Worked examples with the arithmetic reproduced
Note: the following examples are hypothetical and illustrative rather than sourced case studies.
Take a three-book fantasy series at $4.99, $200 spend, $0.40 CPC, 10% conversion, 55% read-through. Clicks are 200 / 0.40 = 500, sales 50, net $3.34. Book-one profit is 50 × $3.34 − $200 = −$33, ROI −16.5%. Add read-through: series value is $3.34 + 0.55×$3.34 + 0.30×$3.34 ≈ $6.18, revenue 50×$6.18 = $309, profit $109, ROI +54.5%. Same spend, counting the books already sold.
A 3-book series at $4.99, $100, $0.35 CPC, 8% conversion, 50% read-through yields 285 clicks, 22 sales, net $3.34. Book-one profit 22×$3.34 − $100 = −$26.52, ROI −26.5%. Series value ≈ $5.85, revenue 22×$5.85 = $128.70, profit $28.70, ROI +28.7% — unprofitable on day two, profitable on day 30.
A 5-book series at $4.99, $100, $0.30 CPC, 10% conversion, 60% read-through yields 333 clicks, 33 sales. Book-one profit 33×$3.34 − $100 = $10.22, ROI +10%. Series value ≈ $7.40, revenue $244.20, profit $144.20, ROI +144%.
A single $2.99 eBook with 0% read-through, $100 at $0.35 CPC, 8% conversion nets about $2.04. 285 clicks, 22 sales, profit 22×$2.04 − $100 = −$55.12, ROI −55%. At $0.99, net $0.35 gives −$92.30. No series rescues it; lower CPC, higher conversion, or a 70% price is required.
When the math says no
Some campaigns genuinely should not run, and the tool is honest about them.
A price below $2.99 with a CPC above $0.20 is hostile. The 35% royalty tier combined with even modest CPC burns margin before read-through can help. Unless book one is a deliberate loss-leader for a higher-priced continuation, the campaign is not viable at that price point. Raise the price into the 70% band or move the ad budget to a different book.
A single standalone book with no series behind it has to be profitable on book one alone unless you have a credible backlist or a newsletter monetization behind the ad. If net is $3.34 and CPC is $0.40 with 8% conversion, cost per sale is $5.00, so each sale loses $1.66. Unless the standalone price is unusually high or conversion is unusually strong, the math says to spend elsewhere. The standalone exception is nonfiction with a high-price backend — a $27 course or $199 workshop can make a negative book-one ACOS profitable on customer value, but that value must be measured and entered as series value, not as hope.
A cost per click that consistently exceeds net per sale even before conversion is a signal that the keyword is too competitive for the price point. If net is $3.34 and CPC is $1.20, you would need a 36% conversion rate to break even on book one, which is not realistic for non-branded search. The calculator flags this as CPC above net, which is the clearest early warning it gives. Change the keyword, tighten the match, or accept that this term belongs to publishers with a longer series behind it.
A read-through rate you have never measured is not a rate. It is a guess that can make any campaign look profitable on paper. Before scaling spend, run a small test at $10 to $20 per day for two weeks, attribute book two and three sales to the same cohort, and compute the observed rate for that traffic source. Replace the guess with that number and re-run the tool. If the observed rate is half your guess, the series value drops by more than you expect.
Series length, pricing ladders, and Kindle Unlimited as a variant
Each additional book adds expected value with diminishing returns from decay. Going from one to two at 50% read-through adds $1.67; from five to six adds about $0.10. Writing book four is strong for ads even though per-reader lift is small, while stretching a weak three-book series to seven for ad math alone is not.
Pricing ladders also matter. Many publishers price book one at $3.99-$4.99 and later books at $5.99-$6.99. Series value is linear in net when royalty is constant, so average net times expected units is accurate within a few percent. Enter the price for the advertised book and treat series value as the blended number.
KU pays per page read, so a long book that is finished is worth more than a short sampled one. A 350-KENP book at $1.50 per full read and 65% completion is worth about $0.97 per borrow. KU read-through is borrows of book two per borrow of book one. The tool uses a $1.00 placeholder; replace it with your KENP-reported dollar-per-borrow after a month.
Budget sizing, daily caps, and the learning window
Start with a $10-$20 daily cap for an exact-match test. At $0.40 CPC, $15 buys about 37 clicks — about 260 per week, enough to see whether conversion is near 8% or 3%. If a keyword hits 100 clicks with zero sales, pause it. If it hits 8-10% near break-even CPC, let it run through a full 30-day read-through window; many series sales from week-one clicks occur in week three.
Scale by multiples, not leaps. A keyword profitable at $15 per day on 37 clicks may not be at $150 on 375 — CPC rises and conversion drops as audience broadens. Increase daily spend 1.5x-2x, observe for a week, then step again. The tool's linear scaling with spend held constant is the optimistic bound; require observed data to confirm.
Include the time cost of the learning window in your decision. Series profit that arrives 30 days after clicks is still profit, but it requires cash flow tolerance. If your budget is $100 total and you need series profit to fund the next $100, you can only spend as fast as read-through monetizes. Monthly budgeting that matches KDP's 60-day payout cycle — spend in month one, collect series revenue through month two, redeploy in month three — is more realistic than daily ROAS targeting for a small publisher.
Tracking read-through from ad cohorts
The only read-through that matters for ad math is ad-attributed read-through, not global read-through.
Pick a clean start date and record daily book-one units from Amazon Attribution or a campaign targeting only the advertised ASIN. Over the same window, record book-two and three units for that cohort. When attribution tags are unavailable, record total daily units for all books and subtract baseline organic sales from a prior no-ad period; the residual is ad lift. Divide lift book-two units by lift book-one units for the cohort to get incremental read-through.
For KU, use daily KENP by title rather than sales; read-through is borrows of book two per borrow of book one for the cohort, and completion matters — a 10% read is not a finished read for long books.
Update the tool after 30 days with the observed ad-attributed rate. The series value that results is the one to budget against. Keep the global rate as a separate note for business planning, but do not use it to justify spend.
How to use the tool in under two minutes
Open the calculator and select the platform your ad actually sends traffic to. Enter the retail price of the book the ad sells, and any print cost if the ad is for the print edition. Set those two once per campaign. Change them only when the offer changes.
Enter your recent CPC for the keyword from the last seven days, not the last hour, and measured click-to-sale conversion for the same targeting. If you have no measured conversion, use 8% for a Sponsored Product exact match as a placeholder and replace it within a week.
Set read-through to your measured ad-attributed rate, or leave it at 50% for a mid-series romance or 30% for a nonfiction sequence as a placeholder, and choose the number of books in the series. Use the number that actually exists and is for sale today, not the number you plan to have next year.
Read the result in order: clicks, sales, net per sale, break-even CPC, then the breakdown of book-one profit versus full-series profit. If book one is negative and series is positive, the campaign is working on the time horizon that matters. If both are negative, the keyword or the price is not viable at that CPC and conversion. If the verdict is pass at the series level, the margin percentage tells you how much room you have to absorb a higher CPC or a lower conversion before the campaign inverts.
Final checks before you scale
Run the tool twice: once with the observed rate and once with rate 10 points lower and CPC $0.10 higher. If the downside stays positive, the campaign is resilient; if it inverts, cap spend and re-measure. The downside bounds attribution noise and CPC variation.
Check that the delivery cost for the file you uploaded matches the assumption behind the net number the tool shows. A 2 MB text-only file and a 12 MB illustrated file at the same price produce different net figures at 70% because delivery is weight-based. Open the KDP pricing tab for the live book and note the listed delivery cost, then confirm the tool's net is within a few cents of that.
Confirm that the series you entered matches the retailer reality. If book three is on pre-order or unavailable in KU, the ad will not generate its read-through for weeks. Count only live, purchasable books in the series length today. When book three goes live, re-run and watch the series value step up.
Then let the window close. Series profit that relies on read-through needs 30 days to be observed. Do not pause a campaign that is break-even on book one and profitable on series after five days because the series bar is still filling. Spend at the rate the downside case tolerates, track the cohort, and scale only when the observed series value confirms the projection.