Writing a Short Ebook People Will Actually Download

Pick the problem you could solve for a friend in one sitting. That is the book.

Short written products fail at the scoping stage far more than at the writing stage. A document that keeps getting closed is a document with no edges, and edges are something you choose rather than discover.

Scoping to one problem so it can be finished this month

Write the outcome in one sentence: after reading this, the reader can do a specific thing. If your sentence contains the word "and", cut one half and keep the other.

Narrow feels embarrassing. Narrow is what gets searched for. "How to get a rental application accepted with no rental history" is a real thing a real person types at midnight. "A guide to renting" is a shelf, and the shelf is already full.

Pick the version of the subject you have actually lived through, not the version you read about. You know the small ugly details: the form with the stupid field on it, the phone menu that hangs up on you, the order you have to do two things in or the second one fails. Those details are the entire reason someone buys your file instead of reading the first search result.

Before you write a word, it is worth checking whether anyone wants this thing, because narrowing the scope is far cheaper before the writing than after it.

The outline that stops it becoming a book you never end

Write the chapter headings as questions a reader would type. Then put them in the order the task actually happens in, rather than the order you find interesting.

Now cap it. Decide the number of headings before you start, write the number down somewhere you will see it, and refuse to add to it. Every extra heading moves the finish line, and the finish line is the thing you are trying to reach.

Under each heading, allow yourself three bullets. If a bullet clearly needs a chapter of its own, it is not a bullet, it is the next product. Put it in a separate file called later and carry on.

The outline doubles as the sanity check on scope. If you cannot see the whole of it on one screen, it is still too big.

Writing sessions: how to structure them so it actually gets done

One heading per session. Sit down knowing which heading, or you will spend the session choosing one.

Draft forward only. No rereading the last paragraph, no fixing the sentence you hate, no reordering. Ugly text can be repaired later. An empty file cannot.

Stop mid-sentence. Ending on a half-finished thought means the next session starts with a rolling start instead of a cold one, which is the difference between writing and staring at a screen.

Keep the later file open beside you. Every tangent, every "I should also cover", every good idea that belongs to a different heading goes in there and stops interrupting the one you are on.

You decided the number of headings, so you know the number of sessions. That count is the honest size of this project, and it plans better than a date does.

Editing your own work when there is nobody to hand it to

Three passes, each with one job. Trying to do all three at once is why editing feels endless.

Pass one is truth. Any claim you cannot stand behind gets cut, or converted into a question the reader can put to someone with actual authority. If you are guessing at a fee, a rule or a waiting time, do not guess in print — tell the reader who to ask and exactly what to ask them.

Pass two is order. Read only the headings and the first line under each. Does step three depend on something you do not explain until step four? Move it.

Pass three is sound. Read it aloud, or paste it into a phone and have the phone read it back. Your ear catches the sentence that doubled back on itself long after your eye stopped seeing it.

Then hand the first two pages to one person who has the problem and ask exactly one thing: where did you get confused? You do not need a copy editor. You need one confused stranger who likes you enough to be honest about it.

Turning the document into a file that opens everywhere

Export to pdf. The layout stays where you put it, it opens on phones without an app, and nobody edits your text by accident.

Set a real document title in the file properties, because that string becomes the browser tab when someone opens it, and "Untitled document" is a poor first impression. Add page numbers. If your tool can build a clickable contents list, build one.

Keep the images small enough that the file downloads on a bad connection in a parking lot, because that is where people open things.

Name the file the way you want it to look sitting in a stranger's downloads folder for years. Not draft-final-3. The actual title, with the version on the end. After that the question becomes where the finished file gets listed, which is a different decision from writing it and a slower one.

The cover, and how much it matters relative to everything else

The cover is a thumbnail. That is the entire brief.

The title has to be legible when the image is the size of a stamp, which rules out thin fonts, low contrast and three lines of subtitle. One idea, big type, a color that is not beige.

Build it in a free design tool with a text block on a solid field, then stop. An hour, not a week. It matters more than the inside of the file at the moment of the click and less than the inside at every moment after, and it is the easiest part of the whole project to hide inside for a two weeks. Cover, title and description work as one unit, and how they fit together is the substance of naming and packaging a digital product.

Medicine, law, tax and the reader who cannot undo it

Some subjects hurt someone if you get them wrong. Medicine, law, tax, immigration, investments — anything where a reader acts on your page and cannot undo it. You can write your own experience of a process: what the form asked, how long the queue was, what you wish you had brought. You cannot write the ruling. Point at the body that can, whether that is the agency itself, legal aid, or a licensed professional.

The second limit is how easily you can be checked. If a reader can verify you in ten seconds and find you wrong, the file costs you more than it earns.

The third is want. Knowing a subject well is a different thing from anyone wanting a document about it. And writing on other people's subjects, for their money instead of yours, is a separate trade with a faster start: freelance writing without any clips pays sooner and teaches you more per hour.

Write about the thing people already come to you about. If nobody has ever asked you anything, pick something else.