Stop Asking What Digital Product To Create. Ask This.

The reason your digital product idea keeps stalling is not that you picked the wrong format. It is that you picked a format at all, before you knew what the buyer needed. Ebook, course, template, planner: every list online tells you to choose from the shelf and start building. I did exactly that. I built the thing first and went looking for someone who wanted it afterwards. It took me an embarrassingly long time to spot what I had inverted, and it comes down to one three-word question I will give you further down.

The format lists are answering a question nobody buys

Search this and you get the same page five times. Ranked lists of product types. Ebooks at the top because they are quick. Then checklists, then Notion templates, then content calendars, then a planner. Some of them add how long each takes to make, or what the author earned last year.

Every one of those lists is supply-side. They answer "what can I make quickly", which is a question about you. Nobody has ever bought anything because it was quick for the seller to produce. They buy because something in their week is annoying, expensive, or slow, and your thing makes it stop.

So the lists are not wrong, they are just out of order. The format is the last decision, not the first. Choosing it first is like buying the packaging before you know what is going in the box. That ordering problem is the same one I unpick in the Practical Income Method, because it shows up at almost every stage of building something people pay for.

I built the thing nobody asked for

Here is my honest low. Early on I decided I was going to make a guide. Not because anyone had asked for a guide. Because a guide felt like the thing you make. I had the tools, I could write it fast, and writing fast felt like progress.

So I had a format and no problem. Which meant I spent days doing the strangest kind of work: reverse-engineering a reason for a document to exist. Broadening the topic so more people might want it. Adding chapters to make it feel substantial. Every one of those decisions made it worse, because I was making it for an imaginary person whose specific pain I had never named.

It went nowhere. Not dramatically. It just sat there, which is how most of these end. And the worst part is that AI made me faster at going nowhere. That is the trap in this whole space right now. You can produce a polished 60-page PDF in an afternoon, so the cost of building the wrong thing has collapsed, and with it the pressure that used to force you to think first. Speed does not fix direction. It just gets you to the wrong place sooner.

The shift: pick the problem, then let the format pick itself

Here is the method. Three steps, in this order, and the order is the whole thing.

Step one: name one specific problem for one specific person

Not "help people with social media". Something you could overhear someone complaining about in real words. "I sit down on a Sunday and I have no idea what to post this week." "I keep rewriting the same onboarding email for every new client." "My product photos look like a phone took them, because a phone did."

Get it that concrete. If you cannot say it in the buyer's own sentence, you do not have a problem yet, you have a category.

The best source is not research, it is memory. What have people actually asked you for help with? What do you get asked twice? Failing that, go and read the complaints: subreddits, Facebook groups, the one-star reviews of tools in your space. People describe their problems in public constantly and in far better detail than any survey would get.

Step two: score the problem on urgency and frequency

Two quick filters before you build anything.

Urgency. Does solving this make or save the buyer money, time, or embarrassment right now? Or is it a someday-nice-to-have? Someday problems get bookmarked, not bought. This is the same instinct behind the classic jobs-to-be-done idea: people hire a product to do a job that is already bothering them.

Frequency. Does this problem come back every week, or did it happen once? Recurring problems justify a bigger price and a deeper product. One-off problems still sell, but they sell as small, cheap, sharp things.

If a problem is low on both, drop it. Do not be clever about it. Go back to step one. This filter alone will save you a fortnight.

Step three: ask what the buyer needs to happen. Learn, do, or done.

This is the three-word question I promised. For your named problem, does the buyer need to learn something, do something faster, or have it done for them? The answer picks your format, and it picks it almost mechanically.

Learn. They do not know how. There is a skill gap and they want to close it. Format: an ebook, a guide, a short email course, a video course. The deeper and more recurring the problem, the more it justifies a course over a PDF.

Do. They know how, they are just slow at it or they start from a blank page every time. Format: a template, a swipe file, a prompt pack, a spreadsheet, a checklist, a Notion system. The product is not knowledge, it is a head start. These are quick to build and quick to prove.

Done. They know how, they do not want to, and they have money. Format: a done-for-you service. Make their videos. Write their emails. Build their product photos. This is where AI earns you the most right now, because the tools let one person deliver at a standard that used to need a small team, and the buyer does not care how it got made. This is genuinely the most under-picked option on every list I have read, because it is not a file, so it does not look like a "digital product". It pays like one.

Run a real example. Problem: "a small skincare brand owner has no product videos and cannot afford a videographer." Urgent, yes, it costs them sales weekly. Recurring, yes, every new product and every campaign. Learn, do, or done? They do not want to learn video. They want it done. So the product is not an ebook about video marketing. It is a monthly video pack. Same person, same starting research, completely different business, and only step three separated them.

Now use AI, and not a minute before

Once the problem, the urgency check and the learn-do-done answer are on the page, AI becomes genuinely brilliant. It will outline the course, draft the guide, build the template, produce the video pack. That is where the leverage lives.

What it will never do is tell you whether the thing is worth making. It has no idea who your buyer is or what is annoying them this week. Feed it a fuzzy brief and it will hand you a confident, well-formatted, perfectly useless product, and it will do it in twenty minutes, which is exactly what makes the mistake so easy to repeat.

So do the thinking on paper. Three lines: the problem in the buyer's words, urgent and recurring yes or no, learn-do-done. Then open the tool. That single ordering is the difference between shipping something and stacking up files nobody downloads.

This works for you, whatever you already know

You do not need an original idea. You do not need a big audience or a niche nobody has touched. You need one problem you have genuinely seen someone struggle with, and the discipline to not open a tool until you have written it down.

And if the answer comes out as "done", start there. It is the fastest paid work available right now for someone with AI skills and no audience, and every delivery teaches you what these people actually struggle with, which is where your ebook or course eventually comes from anyway.

Nobody buys a format. They buy a problem going away.

If you want the full path from a problem to a product someone pays for, my free 5-step guide walks it end to end: finding a problem worth solving, choosing what to build, making it with AI, pricing it, and getting the first sale. It is the same sequence I have laid out here, extended past the choosing stage. Grab the Practical Income Method and start with step one this week.