A physical product has a floor under it. You buy stock, or you buy materials and then spend your evenings turning them into something. Then you buy packaging, and you find out what a padded envelope costs, and you discover that the postage is a business decision rather than an afterthought. None of that is a reason to avoid physical products. It is just money you have to find before you know whether anyone wants the thing.
A digital product does not have that floor. It has a different cost, which is hours of your attention, and hours are not free either. But you can spend them on a Saturday and find out on Sunday whether anyone cares.
Which is why the first product most people can actually finish is a version of something they already know. Not something researched, not something learned from a course about making courses. Something you could explain out loud, to a person, without notes.
What "already know" actually means
It does not mean expert. Expert is a word that stops people from starting.
A more useful test: has anyone asked you the same question twice? Not a stranger on the internet. A friend, a cousin, someone in a comment section, someone at work. Repeated questions are the cheapest market research available, and you have already been given the answers for free.
Some real shapes this takes. Somebody who has been selling on TikTok Shop for six months knows the actual sequence for getting approved, what the samples process feels like, and which three settings cause most of the grief. Somebody who runs a market stall has a spreadsheet that tracks stock and takings and it works, and it took them four attempts to get right. Somebody who sews knows how to resize a pattern for a body that is not the pattern's default, which is a thing thousands of people search for and mostly find bad answers to. Somebody who films in a small flat has worked out where to stand at four in the afternoon so the light is not awful.
None of that requires a qualification. All of it is worth something to a person who is two months behind you.
The one to avoid is the obvious one. If the only thing you know is how to make money online, and you learned it eleven days ago from a video, do not sell that. It is the most crowded category there is, the buyers are the most burned, and you will end up defending claims you cannot support. I wrote about that whole corner of the internet in what works, what doesn't, and what's a scam, and the short version is that the people selling shovels to gold miners have already saturated the shovel market.
Three shapes, and what each one costs in hours
The format is not a branding decision. It is a decision about how much of your life you are prepared to spend before you find out if anyone wants this.
A template is the cheapest thing you can make. A spreadsheet, a Notion page, a Canva set, a Lightroom preset pack, an invoice, a content calendar, a pattern file. Usually it already exists on your machine in scruffy form, and the work is cleaning it up, deleting your own data out of it, and writing instructions that assume the buyer has never seen it before. Call it half a day to two days. The instructions take longer than you think, and they are the part people complain about when they are missing.
The weakness of a template is that it is easy to copy and there is often a free version three search results down. That caps the price. It does not make it worthless. It makes it a five to twenty pound product rather than a fifty pound one.
A written guide sits in the middle. Three to eight thousand words, screenshots, a PDF or a hosted page. Two weekends is a realistic estimate if you already know the content, longer if you keep rewriting the introduction, which everyone does. The advantage is that writing forces you to notice the gaps in what you know. The disadvantage is that a lot of buyers do not read what they buy, and unread products produce refund requests and no testimonials.
A course is the expensive one. Recording, retakes, editing, hosting, the moment three weeks in when you realise module two contradicts module five. Weeks of work, and afterwards a stream of questions from buyers who expect you to answer them, because at course prices they are right to expect that.
There is a cheaper route into the same thing. Sell a live session first. One date, one Zoom link, a fixed number of seats, whatever price feels defensible. You have to prepare, but you do not have to edit, and the questions people ask in the session tell you exactly what the course should contain. Record it. If eleven people paid, you now have both a validated topic and a rough first version. If nobody paid, you have lost a Tuesday evening instead of a month.
Pricing, with the part nobody can tell you
I cannot tell you what your thing is worth. Nobody can, before it has been offered to people who are allowed to say no.
What I can give you is the range people actually charge. Templates tend to sit between five and twenty. Written guides between ten and forty. Courses and cohort things start around fifty and go up a long way, with the higher end usually attached to a name people already trust. Currencies vary, platform fees vary, check what your checkout takes before you set a number.
One thing worth knowing early: a low price does not make the sale easier. A stranger deciding whether to give you nine pounds goes through most of the same steps as one deciding about twenty-nine. They still have to find you, believe you, click through, and type a card number on a phone. You get the same friction for a third of the money. So do not price at three pounds because you are nervous. Price it at the low end of the range and see what happens.
The arithmetic afterwards is unforgiving in a boring way. If two per cent of the people who reach your checkout buy, and the product is fifteen, then a hundred people arriving at the checkout is about thirty pounds. Everything upstream of that has to be big enough to produce a hundred arrivals. I went through those numbers properly in how long until the first sale, and the short version is that the machine needs feeding before it will give you an opinion.
The bit between the video and the checkout
Here is where most first digital products quietly leak. You end up with more than one thing. The free sample, the paid guide, the template bundle, the waitlist for the live session. And one link in a bio.
This is the problem Relay is for. It is a link-in-bio page, made in Warsaw. Unlimited links on every plan, including the free one, which needs no card and does not expire. You get your own photo and bio, social icons, five themes, and headings and dividers, which matter more than they sound. A heading that says "free stuff" above two links and another that says "paid" above one is the difference between a page and a list. There are tap and view counts, so after a fortnight you can see which of your three products people actually reach for, rather than which one you enjoyed making. There is a QR code and a story-sized share image for the places where a link is not clickable.
Relay is also, being fair to the problem, part of it. We do not take payments. There is no checkout, no cart, no digital delivery. Your buyer taps a link on your Relay page and lands on Gumroad or Payhip or Lemon Squeezy or Shopify or wherever you set up, and that hop across to a different domain loses you some people. It always does. We also have no custom domains, so the page is relayme.bio slash your name rather than your own address. If those two things are dealbreakers, they are dealbreakers, and we say so on the comparison page, which includes the rows where Linktree beats us.
What we do instead is the page itself. Pro is thirty dollars a year, or four dollars a month, plus VAT, and adds all forty-seven themes, all eight fonts, custom colours, a background image, per-link images, scheduled links for a launch date, inline players, and email capture. Email capture is the one that matters for a digital product, because the people who are not ready to buy today are the only free audience you own. Data sits in the European Union, fonts are self-hosted, there is no advertising and no third-party analytics on your page.
What it will not do
A digital product does not create demand. It removes the stock problem and leaves every other problem intact. You still need people to see something, care, and click, and the thing that gets them there is the same short video or the same small ad budget it would have been for candles.
It is not passive. The file is passive. The selling is a job you do most days.
Somebody will share your PDF in a group chat. This is not preventable in any way worth the effort, and building a fortress around a twelve pound guide costs more than the leaked copies do. Price with that assumed.
Refunds happen, support emails happen, and people will ask you questions the guide answers on page four. Some buyers will be disappointed. A few will say so publicly. That is a normal cost of selling anything to strangers, and it does not mean the product is bad.
And it will not survive being invisible. If you are stuck before the first sale rather than after it, the product is probably not the bottleneck, and this post on the first sale is more useful than another week of editing.
So: pick the question you have answered twice. Write the smallest useful version of the answer this weekend. Put it behind a checkout, put the checkout behind one link, and put the link somewhere people can find it. Then go and make the thing that sends people there, which was always the actual work.