In short: The digital products worth selling solve a specific, recurring problem for a defined buyer: templates, spreadsheets, practical guides, courses, data sets and small software tools. Generic printables and AI-written ebooks are easy to make, which is why they are crowded. Choose something that draws on knowledge a buyer can’t quickly get elsewhere.
The attraction is simple. Once built, a digital file can be sold repeatedly with no stock and no shipping. The catch is that low barriers apply to everyone. Generative AI has made it easier still to fill a marketplace with passable files, which makes “easy to produce” a warning sign. What holds value is the part that is hard to copy: experience, a tested process, or data that took effort to gather.
The main categories compared
| Product | Good example | Effort to build | Typical challenge |
|---|---|---|---|
| Templates | Client onboarding kit for wedding photographers | Low to medium | Easy to copy; buyers compare on price |
| Spreadsheets and calculators | Job-costing workbook for small contractors | Medium | Needs real domain knowledge and support |
| Practical guides | Step-by-step guide to a specific permit or exam | Medium | Must stay current |
| Online courses | Skills training with exercises and feedback | High | High refund rate if outcomes aren’t clear |
| Stock assets | Fonts, sound effects, illustrations, photo presets | Medium to high | Saturated marketplaces and AI competition |
| Data sets and directories | Updated list of grants for a specific industry | Medium, then ongoing | Requires regular updates |
| Software and plugins | Small tool for one workflow | High | Bugs, support, compatibility |
| Printables | Planners, worksheets, wall art | Low | Very crowded and low priced |
The table suggests a useful pattern. Products aimed at people who use them for work can usually carry higher prices than those aimed at hobbyists. A $79 workbook that saves a contractor hours on every quote is easy to justify. A $79 wedding planner printable isn’t.
What makes an idea worth building
Before spending weeks on anything, run the idea past these questions:
- Who exactly buys it? “Small business owners” is too wide. “Independent bookkeepers who take on restaurant clients” is a group you can find and speak to.
- What happens if they don’t buy? Money lost, hours wasted or a failed exam can all justify a price. Mild inconvenience usually doesn’t.
- Why you? Your job, a problem you’ve solved or research you’ve done is what separates the product from a free search result.
- Can they get it free? If a chatbot produces something roughly as good in one prompt, buyers need a strong reason to pay. A tested method, specific examples and maintained data all help.
- Can you reach buyers? A good file with no audience sells nothing. Our guide on how to sell digital products covers the route to a first sale.
If the idea survives, check demand with real money before you build it all. How to validate a business idea sets out cheap tests, including pre-selling.
Why the price and platform matter
Fees take a bigger share of cheap products, and this is often overlooked. Here is what remains from a $25 sale and a $5 sale, using published fee rates and our own arithmetic. These are US figures and they can change, so check before you commit.
| Sale | Etsy fees (6.5% + 3% + $0.25 + $0.20) | Gumroad direct (10% + $0.50, plus about 2.9% + $0.30) |
|---|---|---|
| $25 sale | $2.83, keep $22.17 | About $4.03, keep about $20.97 |
| $5 sale | $0.93, keep $4.07 | About $1.45, keep about $3.55 |
Marketplace rules add variations. Etsy’s optional Offsite Ads charge of 15% or 12% applies to orders from those ads, and is mandatory once a shop passes $10,000 in yearly sales. Gumroad charges 30% on sales it generates through its own Discover marketplace. On the other hand, Gumroad now acts as merchant of record, so it collects and remits sales taxes and VAT, which is a real administrative saving.
Small prices also create a volume problem. On the Etsy figures above, a $5 file needs about 250 sales to bring in $1,000. A $79 product needs about 15, which also means fewer customers to support. Lower prices are not automatically a mistake, but they have to be matched by much larger traffic.
Ideas we would skip
Some products are heavily promoted because they are simple to teach, and they tend to be poor businesses for a newcomer:
- Mass-produced AI ebooks. They are easy to generate and hard to differentiate, and buyers are increasingly wary.
- Generic planners and journals. They compete with a huge supply on platforms where established sellers have years of reviews.
- Resale rights bundles. Buying a pack you can resell means you and thousands of others hold the same files. Check the license too, because rights often limit what you can do.
- Courses on making money from courses. This is a thin market with a credibility problem.
There is also a question of honesty. A product should do what its page says. Inflated income claims are a matter for regulators, and they create refunds as well.
Starting points that often work
Look at your paid work first. What did you build for yourself that colleagues asked to copy? Which questions do newcomers to your field keep asking? What spreadsheet or checklist has quietly saved you time for years? Those items have been tested in use, and a buyer who shares your job will understand the value quickly.
Selling files can also sit alongside other activities. A freelancer who knows what clients will pay can package repeat work into a kit for those who can’t afford the full service. If you are still choosing between models, read our overview of how to start an online business.
Frequently asked questions
Can I sell products I made with AI?
Usually, but check the platform’s rules and disclose where required. Copyright protection for purely AI-generated material is limited in some places, including the US, so competitors may be able to copy it. The bigger issue is commercial: buyers pay for quality and judgment, so human testing and editing should be a meaningful part of the product.
How much can a single digital product earn?
There is no reliable average, and income screenshots online are not a sound guide. Results depend far more on audience and problem than on the file. Plan with your own numbers: visitors you can realistically reach, a cautious conversion rate of 1% or 2%, price, then fees. If the total is disappointing, change the plan.
Should I offer refunds on digital products?
A clear refund policy usually helps sales and reduces payment disputes, even though a buyer could keep the file. Consumer law may also set minimum rights, and these differ by country. For courses or tools, a time-limited guarantee is common. High refund rates are useful feedback that the sales page and the product don’t match.
