Digital Products to Sell, and What They Actually Net

Every list of digital products to sell that currently ranks for this phrase was written by a company that wants you to sell on their platform. Whop, Wix, Printful, Payhip, Skillshare. I checked. One of them promises 114 ideas, which is roughly 113 more than anyone needs, and none of them tell you what you keep after fees.

So this is the other kind of post. Fewer ideas, more arithmetic.

Short answer for the skimmers. The digital products that sell most reliably are the boring functional ones, templates and planners and presets, because they solve a problem the buyer can name before they arrive. What you keep depends less on your price than on where you sell, and the spread is wider than people expect. On a $29 product you might keep $28.42 or you might keep $20.30, same product, same buyer, different checkout page. I'll show that math below.

What Actually Sells, In Order Of How Sure I Am

Templates first. Notion setups, spreadsheet systems, Canva packs. Someone searching for a Notion template has already decided they want one, which is most of the sale done before you show up.

Planners and calendars next, including the digital planner category and the digital calendar wall variants people set up on tablets. Seasonal, but it's a real season.

Then creative assets. Lightroom presets, fonts, mockups, stock photo bundles. Lower prices, higher volume, more competition from free.

Then educational stuff. Ebooks, courses, toolkits. This is where I actually live, and I'll be honest that it took me longer to get traction here than the templates people seem to get.

Print on demand sits slightly outside all this because you're selling a physical object with a digital file behind it, so the margins work completely differently and the fees below don't apply the same way.

I'm not going to pad this to 114. If you can't picture the buyer, the idea isn't ready.

A note on competition, since the lists never mention it. These categories are not equally crowded, and the crowding doesn't track with the volume numbers. Presets and fonts look attractive because the search demand is enormous, but you're competing against a wall of free alternatives and against people who've been building preset packs since before it was a category, so the realistic outcome for a new seller is a lot of listings and very little movement. Templates for a specific tool are friendlier, partly because the buyer pool renews constantly as new people adopt the tool. Planners are seasonal in a way that catches people out, with a spike around the new year that flatters your first month and then leaves.

What A Product Should Cost

I've priced things between $29 and $49 and I still don't have a confident theory, which I'd rather admit than dress up.

What I've noticed is that the price mostly changes who complains. Cheap products attract buyers who want support out of proportion to what they paid, and expensive ones attract buyers who want the thing to be transformative rather than useful. Somewhere in the middle you get people who bought a tool, used the tool, and left you alone, and honestly that's the customer you want when you're one person.

The fee math nudges this too, and not in the direction people assume. Gumroad's flat fifty cents per transaction barely registers on a $49 sale and takes a real bite out of a $5 one. If you're thinking about a cheap high-volume product, run the per-unit numbers before you commit, because the percentage isn't the whole fee.

Test two prices if you can stand it. Not on the same day, and not with elaborate statistics, just run one for a few weeks and then the other and see whether anything moves. The result usually won't be dramatic. That's information too.

The Fee Math Nobody Shows You

Here's the part the platform listicles skip, for obvious reasons.

I pulled these numbers on 28 July 2026 from the platforms' own pricing pages. Gumroad charges 10% plus $0.50 per transaction on sales that come through your own profile or a direct link you send. If a new customer finds you through their discover marketplace instead, it's 30%. Payhip runs three tiers, free at $0 a month with a 5% transaction fee, Plus at $29 a month with 2%, and Pro at $99 a month with no transaction fee at all. On Payhip, Stripe and PayPal still take their own cut on top, which is easy to forget when you're comparing the headline percentages.

I wanted to check Etsy against these and couldn't. Their fee page returned a 403 to me both times I tried today, so I'm not going to quote you a number I didn't personally load. Go read it yourself before you commit to that channel. I'd rather leave a hole here than fill it with something I half-remember.

On a single $29 sale, before Stripe or PayPal touch it:

That $20.30 row is the one worth staring at. Same product, same price, and the difference between someone finding you through a marketplace versus clicking your own link is more than eight dollars. Sell a hundred of those and you're arguing about eight hundred dollars, which is a real amount of money.

The monthly plans change the picture once volume shows up, though, and this is where people pick wrong.

When The Paid Tiers Start Winning

Payhip Plus costs $29 a month and saves you 3 percentage points versus Free. So it pays for itself once 3% of your monthly revenue exceeds $29, which is revenue of about $967. Call it a grand.

Pro costs $99 and saves you the full 5 points against Free, so it needs 5% of revenue to clear $99. That's $1,980 a month.

Below a thousand a month the free tier wins and it isn't close. Most people reading a post about digital products to sell are below a thousand a month. Some of them will still subscribe to Pro in month one because it feels like the professional choice, and they'll pay $99 for the privilege of keeping less money.

I want to be careful here. These are list-price comparisons, not a recommendation of one platform, and payment processing sits on top of all of them.

My Own Numbers, Including The Bad Ones

I've shipped seven toolkits priced between $29 and $49. Across everything I've sold, all products, all time, it's somewhere around ten thousand dollars. Lifetime. Not monthly, not recurring, lifetime, and I mention that because the number gets quoted at me occasionally in a way that suggests people think it's a monthly figure.

The first sale on one of my sites was $29.99. It landed roughly two months after launch. I'd been refreshing the dashboard for most of those two months like refreshing would do something.

Here's the more useful failure. That same site sat at an average position of 47 in Search Console for a long stretch. Position 47 means nobody has ever seen it. You can have a perfectly good product, a working checkout, decent copy, and still make nothing, because the product was never the bottleneck. Distribution was. I kept improving the product anyway, because improving the product is comfortable and doing outreach isn't.

If I could hand my earlier self one line it'd be that. Not "make a better template."

How Much Can You Actually Make

People ask this constantly and the honest answer is unsatisfying, so most posts replace it with a screenshot of somebody's best month.

What I can tell you is the shape of it rather than the number. The first stretch pays nothing at all, and it lasts longer than you're budgeting for. Mine was about two months to a single sale, and I'd built the thing properly, with a working checkout and copy I'd rewritten three times. Then it goes lumpy. You get a sale, then eleven days of nothing, then two in an afternoon, and you start building theories about why that have no basis in anything. Then, if the distribution work is happening in the background, it turns into something that resembles a trickle rather than a series of events.

Ten thousand lifetime across seven toolkits and a pile of other things is not a story about a great month. It's a story about a lot of months, most of them small, and I think that's actually the normal shape for anyone who isn't already sitting on an audience when they start. If you already have an audience, ignore most of this post, your economics are completely different and better.

The thing I'd push back on hardest is the framing where you pick the "most profitable" category and that decides your outcome. Category matters much less than whether anyone knows the product exists.

The Questions That Come Up Every Time

Do digital products still work in 2026, given that AI can generate a lot of this? Yes, but the floor fell out of the generic end. A pack of generic presets or a bland Canva template competes with free and with anyone who can prompt their way to something similar in a few minutes. What holds value is specificity, a template built around a workflow you actually understand, with the opinionated defaults already filled in. The buyer isn't paying for the file, they're paying to skip the thinking.

What sells best for someone starting out? Templates, almost always, because the buyer arrives already knowing what they want and you don't have to teach them why they need it. Courses are the opposite. You have to convince someone the problem is worth solving, convince them you can solve it, and then convince them of the price, and that's three arguments instead of zero.

Do you need your own site? Not at first. A direct link into a Gumroad or Payhip checkout is a perfectly good business for a long time, and the fee table above says the direct link is where your margin lives anyway. I'd get the product selling somewhere before I spent a weekend on a storefront, which is advice I did not personally follow.

What about refunds? Budget for some. Digital goods get refunded more casually than physical ones because there's nothing to post back, and a small refund rate is just a cost of doing this rather than a sign something's broken. If it's large, that's a product problem or a promise problem, and it's usually the promise.

The Part About Marginal Cost

One genuinely nice thing about digital products, and it's underrated in these lists. Once the file exists, the hundredth copy costs you nothing. I generate a lot of my own assets locally on an M4 Pro, so the marginal cost of another variation is basically electricity and my own time.

That changes what a sensible product looks like. If each unit costs nothing, a $12 product that sells steadily can beat a $97 product that sells twice, and you don't have to agonise about which is more "premium". Test both. The test is cheap.

Physical goods don't let you think that way. That's really the whole advantage, and it's worth more than any specific idea on any of those listicles.

Where The First Buyers Come From

This is the section I'd have skipped two years ago, so I'm putting it near the end where the people who are still reading will find it.

Search does not work at the start. Not because SEO is dead, but because a new domain sits outside the range where anyone sees it, and the gap between "indexed" and "visible" is enormous. My position-47 stretch is the clean example. The pages were indexed. Google knew they existed. Nobody was ever going to scroll to them, and no amount of editing the product changed that, because the product was never what was being tested.

So the first buyers come from somewhere else, and every option is some flavour of talking to people directly. Communities where your buyer already gathers, if you've been a real participant there and aren't just arriving with a link. Direct outreach to individuals who have publicly described the exact problem your thing solves. A small list you've been building. Whatever platform you personally have any presence on, even a bad one, because a channel with seven subscribers still converts better than a search result nobody can reach.

Seven subscribers. I'm using that number because it was mine, and because it sounds like a joke until you compare it against position 47, where the effective audience is zero.

The uncomfortable part is that this work doesn't compound the way product work feels like it does. You do it, you get some sales, and then next week you do it again. It only starts compounding once search finally catches up, which for a new site is months out, and the whole game in the first stretch is staying alive long enough to reach that.

What I'd tell someone starting today is to decide the distribution channel before the product. Not after, not "once it's ready". If you can't name the specific place you'll tell the first fifty people about it, that's the thing to fix, and it's fixable in an afternoon of honest thinking, whereas discovering it four months later costs you four months.

What I'd Do With A Blank Page

Pick one template-shaped product for an audience you can already reach. Sell it from a direct link, not a marketplace, so you keep the 10% version rather than the 30% version. Stay on a free tier until your own numbers say otherwise, which for most people means until you're clearing a thousand a month.

Then spend the time you would have spent making product number two on getting anyone at all to see product number one.

I know how that sounds coming from a post that just listed product categories at you. The categories are the easy part. They were always the easy part.