Digital Product Pricing Strategy: Net Math, Tests, FTC Rules

You have a digital product ready and a price box that is still empty, and most advice on filling it stops at telling you to charge what it is worth. This guide gets you to an actual number. It shows what each common price nets after Stripe, PayPal, Gumroad, Lemon Squeezy or Payhip fees, what the FTC says about “was” prices, and a test template with a stop rule you can hold yourself to. By the end you should have a starting price and a written plan for checking it.

This is not licensed legal, tax or financial advice. Fees change, so recheck the pricing pages in the sources before you commit. There are no affiliate links here.

To see the net at any price, use the digital product fee calculator.

What each price leaves you after platform fees

The sticker price is what the buyer pays. The number that matters for every decision below is what you keep.

The table shows what one sale nets at five common price points, with the share taken in parentheses. Rates come from each company’s official page, fetched September 28, 2026, and a script did the arithmetic in whole cents. All are as of September 2026.

Read the Gumroad row as a best case. Its help center says the 10% + $0.50 does not include card processing, which it lists as 2.9% + $0.30, so a card sale nets less: $7.04 at $9 and $41.88 at $49.

Two patterns decide a lot. At $9 the fixed part of each fee does most of the damage: Gumroad’s 50 cents alone is 5.6% of the price. By $49 the percentage takes over and the rows settle into their long-run order.

The other pattern is what the extra percentage buys. Lemon Squeezy, Gumroad since January 1, 2025, and Stripe’s Managed Payments act as merchant of record, which means they handle sales tax and VAT on your sales. Plain Stripe and PayPal leave that to you. Payhip sits in between: since July 1, 2026 it collects US sales tax and Canadian GST/HST/PST, alongside the EU and UK VAT it already handled. If most of your buyers are outside the US, part of that gap is a service you would otherwise pay for in hours. If they are nearly all domestic, it is mostly margin.

For the processor-level detail at other price points, including disputes and refunds, see Stripe vs PayPal fees for digital products. The platform comparison is in Gumroad vs Lemon Squeezy vs Payhip.

Set the starting number: value ceiling, net floor, competitor band

Three numbers box in a sensible first price. You need all three written down before you type anything into the price field.

The ceiling is what the result is worth to the buyer. Write the outcome in one line, then ask what the buyer would spend to get it another way: hiring someone, buying a pricier course, or spending their own weekends. Your price should sit well under that, so the purchase feels obvious.

The floor comes from your fixed costs and the net column above, not the sticker price. Say your email tool, hosting and design subscriptions total $45 a month (a made-up figure) and you honestly expect 10 sales a month. Each sale has to net at least $4.50 before you earn anything. That rules out very little at $19. It rules out a lot of $5 ideas once fees are in.

The band is what buyers already see. Open five products that solve the same problem in the same format and note their prices. Landing inside that range needs no explanation on your sales page. Landing above it needs a reason the buyer can see in ten seconds.

My position: pick a number in the upper half of the band that clears the floor with room to spare. You can test downward later, and a lower price is easier to justify than a higher one.

One price, tiers, or a bundle

Start with a single price unless you already have the extra material a second tier would need. Tiers add decisions for the buyer and support work for you.

When you do add structure, bundling is where fee math helps most. Three separate $9 items through Stripe net $25.32 together. One $27 bundle of the same three nets $25.92, because you pay the 30-cent fixed fee once instead of three times. Through Lemon Squeezy the difference is a full dollar, $24.15 against $25.15, and through Gumroad it is $22.80 against $23.80.

That dollar is small. What matters more is that a bundle moves you out of the price range where fixed fees hurt most.

For tiers, the common shape is a core product, the core plus working files, and a top tier with something that costs you time, like a review call. Keep the top tier only if you actually want to deliver it at that price. A tier built only to make the middle look cheap still has to be honored when someone buys it.

Charm prices like $27 instead of $30 are cheap to try. I have not found a primary source for the effect sizes that get quoted for them, so treat a charm price as one more variant to test rather than a known win.

Launch discounts and “was” prices: what the FTC guides require

Launch pricing is where honest sellers get into trouble by accident. A strikethrough price on a product that has never been sold at that price is a former-price claim, and the FTC has written guidance on exactly that.

Checklist, based on the FTC’s Guides Against Deceptive Pricing at 16 CFR Part 233, read on eCFR September 28, 2026:

1. Show a “was” price only if you actually offered the product at that price, openly, in good faith, for a reasonably substantial period, in the recent regular course of business.
2. Don’t write “formerly sold at” unless real sales happened at that price. An offered price doesn’t need sales to count as a former price, but wording that implies sales does.
3. Never raise a price briefly so you can advertise the drop back to normal. The rule’s own example is a retailer inflating $7.50 to $10 for a few days and then advertising $7.50 as a bargain.
4. Keep reductions meaningful. The guides call a “reduced to $9.99” claim from $10 misleading.
5. If you say an offer is limited, limit it. The guides say sellers should not make a “limited” offer that in fact is not limited.

For a launch, that points to one safe pattern. Skip the strikethrough and state the future: “Launch price $19 until October 15, then $27.” Then raise it on October 15. After the product has sold at $27 for a real stretch of time, a later sale back to $19 can honestly reference $27.

These are federal guides, and state consumer-protection rules can add requirements. If discounts are central to your business, ask a lawyer to look at your sales page once.

How to test a price without fooling yourself

Most price “tests” by solo sellers are two weeks at one price, two weeks at another, and a feeling. The template below adds the three things that make the result mean something: a hypothesis written in advance, a stop rule, and a log of everything else that changed.

Price-test template. Copy it into a doc before you change anything.

• Hypothesis: “Raising the price from $19 to $27 will keep net revenue per sales-page visitor at or above the current level.”
• Variants: A = current price, B = new price. One product, one sales page, nothing else changes.
• Schedule: alternate by week (A, B, A, B) rather than one long block each, so one good week can’t decide it. Everyone who visits in a given week sees the same price.
• Stop rule: stop when each price has the number of sales in the sample-size table below, or after eight weeks, whichever comes first. Decide which before starting.
• Decision metric: net revenue per visitor = (sales x net per sale from the fee table) / unique sales-page visitors. Not conversion rate, not gross revenue.
• Log daily: unique sales-page visitors, checkout starts, sales, refunds, net after fees, traffic sources, and any email send, ad spend change, launch or outside mention.
• Void the week if: you changed ad spend, sent a launch email, or got a spike from an outside mention during it.

The break-even tells you how much conversion can fall before a higher price loses money. Computed on Stripe net per sale from the table:

So the $19-to-$27 test is asking one question: did conversion fall by more or less than 30%?

Here is the part that tends to get skipped. How many sales it takes to answer that depends mostly on the size of the drop, not on your traffic or conversion rate. Using the standard two-proportion formula at 95% confidence and 80% power:

If you sell a few dozen copies a month, you cannot measure a 30% difference in eight weeks. That’s fine. It means the test gives you a direction, and you should decide by rule before seeing the data. Mine: keep the higher price if net revenue per visitor at the new price is at least equal, and go back if it is clearly lower.

The void rule comes from my own ebook. It sold while I paid for ads, and when I stopped the ads the sales stopped, which is why I eventually shut it down. Had I tested a price during a month when ad spend changed, I would have measured the ads and credited the price.

When price is not the problem

A price test assumes people reach the page. If sales-page visitors dropped along with sales, changing the price just sells the same missing traffic for a different amount.

Check the visitor count before you touch the price. If traffic fell, the fixes are in the five-cause diagnostic for an ebook that stopped selling and in selling an ebook without ads, not here. If traffic is steady and checkouts fell, then price, offer and page copy are fair suspects, one at a time.

Raise the price when three things are true: traffic is steady without paid spikes, the higher price clears the break-even in your log, and you can give buyers a dated notice. More launch planning lives in the launch and growth guides.

Today, take your current or planned price, find its net in the fee table for your platform, and write the one-line hypothesis and stop rule from the template before you change anything.