A long explainer article is not a long story. It is a chain of functional blocks, each doing one job: hook, lede, nut graf, background, core sections, evidence, limits, takeaways. If you want to know how to structure a long explainer article, write the sentence you want the reader able to repeat after the first 200 words, then build every section backwards from it.
One clarification before we start. Most results for this question are about explainer videos — 60-second scripts, storyboards, hook-first structures for a moving image. This is about the written explainer: the 1,500 to 3,500 word article that has to carry a reader through a process, a policy or a piece of science they have no background in. The two formats share almost nothing structurally, and mixing their rules is the most common reason an explainer reads like a checklist.
Length is not the hard part. Plenty of 3,000-word explainers lose readers in paragraph four, and plenty of 1,800-word ones hold them to the end. The difference is almost always the same thing: the writer front-loaded the wrong material, then tried to repair it with subheadings.
What You Need
Before you open a draft, you need seven things. If any one of them is missing, the structure will have to carry weight it cannot bear.
- A defined reader question. One sentence, in plain words, describing what a reader will be able to do or understand when they finish.
- Verified reporting. Interviews, documents, datasets and observations you can attribute. Explainer structure amplifies whatever you put in it, including the parts that turn out to be wrong.
- An evidence inventory. A list, not a pile: every fact, number, quotation and counterargument, each tagged with the section it might serve.
- A proposed narrative outline. The order of the sections, written before any prose. One line per section is enough.
- Section headings in reader language. Written as the question the section answers, not as a label for the reporting you did.
- Source notes. Wherever a number came from, recorded at the moment you used it. Structure work is renumbering work, and you cannot renumber what you cannot find.
- A fact-checking method. Decide in advance who checks which claims, and when. A named checker beats a general promise to be accurate.
Budget about an hour for the outline and a similar block for the first draft of the first two sections. The whole pre-writing stage is two to three hours, and it saves days of rewriting a piece that was shaped wrong from the start.
Step-by-Step: How to Structure a Long Explainer Article
Seven stages, run in order. The point of running them in order is that each stage produces a usable part of the outline, so you are never drafting blind. If you cannot finish a stage, the article is not ready to be written yet.
1. Define the reader’s central question
Turn the assignment into a single question the article will answer, and only that question. “Cover the new data-sharing rules” is an assignment. “A reader who runs a small clinic needs to know what changes for them on 1 March and what they must do about it” is a question that can structure a piece.
Name the reader too, because the same topic produces different explainers for a general audience, a professional audience and a specialist audience. The vocabulary you may skip changes completely between those three.
Then set scope. Write down what the article will not cover, and put that list in your notes. Most abandoned explainers were not too short; they carried a scope nobody agreed to.
2. Build an evidence inventory
List everything you have: reported interviews, official documents, datasets, prior reporting, academic work, the counterargument from someone who disagrees, and your own observations. Tag each item with the section it could plausibly support.
Then mark the gaps. Anything tagged “somewhere” is a hole, and it is usually a hole in the reader’s vocabulary rather than in your reporting. Those are the questions you go back out and answer before drafting.
Decide now what to cut, because the material you researched and did not use will follow you into the draft otherwise. A piece earns its place by answering the central question or supplying a term the reader needs. Everything else belongs in your notes, and maybe in a second piece.
3. Choose a narrative progression
There are four progressions that work for written explainers, and mixing two of them is the most reliable way to lose a reader.
Problem, then solution. Open on the thing that is broken or newly possible, then work through the cause and the fix. Best for policy and service journalism, where the reader arrived with a practical need.
Cause and effect. Start at the origin and follow the chain forward to where it lands today. Best for investigations and history-driven science writing, where the reader genuinely does not know where the story begins.
Question-led. Each section answers one question the reader is actually holding, in the order a curious person would ask them. Best for evergreen explainers that people will arrive on from search.
Chronological. Use it only when sequence is the explanation itself, such as a walkthrough of how a rule was written. Anything else and readers lose the reason for reading.
Pick one and write its name at the top of the outline. The KSJ Handbook’s advice on choosing a story treatment is the same move at the assignment stage: decide what shape the reader needs before writing a word of prose.
4. Map sections to real reader questions
Turn your central question into five to nine sections. Each gets an H2 phrased the way a reader would phrase it, and each H2 should be answerable in two or three sentences. If a heading cannot be answered briefly, it is two sections wearing a trench coat.
Place the evidence under the question it supports, not under the topic it belongs to. This is the single most effective restructuring move, and it regularly moves two or three paragraphs into entirely different sections.
Give each block a word budget rather than a total. The table below is what I hand to writers on a 2,500-word explainer; the numbers are a starting shape, not a rule.
| Block | Job it does | Budget on a 2,500-word piece |
|---|---|---|
| Hook and headline | Gives the reader a reason to stay | 15–25 words |
| Lede | States what this is and what happened | 40–60 words |
| Nut graf | Explains why it matters to this reader | 50–80 words |
| Background and context | Builds the vocabulary the core needs | 15% of total |
| Core sections | Delivers the explanation in dependency order | 50–60% of total |
| Evidence and data | Proves the claims already made | 15–20% of total |
| Limits and counterargument | Shows where the explanation stops | 8–12% of total |
| Takeaways | Gives the reader something to do or remember | 60–100 words |
Now prune. Any section that repeats context another section already supplied goes. Any section that does not advance the central question goes to a sidebar, a linked piece or the bin.
5. Draft layered sections
Every section gets three layers. Open with the main point in one or two sentences, so a reader who skims still gets the argument. Develop it with the evidence, the example and the context that point needs. Close it with a transition that says why the next section exists.
Keep paragraphs to one or three sentences and one idea each. On a screen, a paragraph of eight sentences reads as a wall, and readers abandon walls.
Write in dependency order: each section may assume only what earlier sections have already explained. When you catch yourself using a term a section introduces after the section that needed it, the outline is out of order, not the prose.
Put the caveats with the claim they limit, not in a block at the end. A reader who has just absorbed an argument deserves its weakness immediately, while the argument is still in mind.
6. Add visual support selectively
A visual earns its place when it shows a relationship the sentences cannot: a timeline, a before-and-after comparison, an annotated chart, a flow of steps, a diagram of a process. If the paragraph already communicates it clearly, the image is decoration and it costs you load time and attention.
Use sidebars and pull quotes as overflow containers rather than as emphasis. The KSJ Handbook treats this as the pressure valve for information that does not fit the narrative, and it is the right use: a reader who wants the detail finds it without anyone else losing the thread.
One caution from the interactive side of the craft. The Explaain team put a finger on the failure mode that haunts long explainers, describing the reaction readers have to being taken through a walkthrough tutorial rather than given an answer. Keep that in mind before you decide to break your explainer into six scroll-driven panels. Interaction is a cost in comprehension, and you pay it only when the format is doing work the prose cannot.
Jargon has a readability rule worth more than most formatting advice: define a term the first time it appears, and introduce no more than one unfamiliar term per sentence. A reader who has to hold three new definitions at once will drop the piece and blame themselves.
7. Edit for sequence, clarity and retention
Edit in three passes, in this order, because each pass tests a different thing.
Sequence. Read only the headings, in order, with the prose hidden. Do they tell a coherent story? If two headings could swap without anyone noticing, the order is arbitrary and readers will feel the seam.
Clarity. Cut every sentence that repeats something already stated. Move anything you cut-and-pasted twice into a sidebar or delete it. Then check that no section depends on a term the reader has not met.
Retention. Find the point where you would stop reading as a reader, usually between 1,200 and 1,800 words, and put a reason to continue there: a question the next section answers, a chart that resolves something, a short section that reads in under a minute. Vary the length and shape of your sections deliberately. Five identical 400-word blocks are harder to read than three varied ones, even when the total is the same.
Check the mobile view last, because that is where a long piece is actually read. One idea per paragraph, generous spacing, subheads that survive a narrow column, tables that scroll rather than squash, and alt text that describes what the image shows.
Common Mistakes
These are the failures I see most often, each with the fix that usually works.
Burying the lede. The first 200 words go to scene-setting, a quotation about mood, or three paragraphs of history. Fix: write the lede last, not first, and paste it in at the top afterwards. It is much harder to bury a paragraph you wrote deliberately.
Reporting every finding. Everything you researched makes the piece, so the piece has no shape. Fix: apply the central-question test from stage two. Evidence that neither answers the question nor supplies a needed term does not ship.
Using one total word count. The writer gets “2,000 words” and spends it on background, then rushes the explanation. Fix: budget per block, as in the table above, and treat the total as an outcome rather than a target.
Buried main points. Each section opens with throat-clearing before the point arrives. Fix: first sentence is the point. A reader who reads only the first sentence of every section should still come away with the argument.
Too many subsections. Eight H3s under one H2 signal that the section has not decided what it is. Fix: if two headings need the same evidence, merge them.
Repeated context. The same background arrives in three different sections because each was drafted without reading the others. Fix: assign every fact one home section during outlining, and cross-reference instead of repeating.
Transitions that assert instead of connect. “Now let’s look at the second factor” tells the reader the section exists rather than why. Fix: each transition should carry a piece of logic, not a signpost.
Overlong paragraphs. Fourteen-line blocks read as a wall on every device. Fix: one idea per paragraph, three sentences maximum, cut the connective tissue.
Visuals added for decoration. Stock imagery and charts that restate the sentence beside them. Fix: every visual must carry something the prose cannot say in the same space.
Format inertia. The same shape gets used for every story because the section or the column already exists, so the writer fills in blanks rather than serving the reader. Smartocto names this and its sibling, format mixing, as the two real failure modes in format choice, and the fix is the three-question test it proposes: topic, user need, format, in that order. Fix: decide the format per piece, and if a piece is part report, part analysis and part reaction, split it.
Two habits that prevent most of the above. Write the takeaway sentence before you write the first section, and put it at the end to see whether you got there honestly. Then keep a “what this does not explain” note in your source file, and mine it every time you rewrite — that material is usually the strongest section you are not writing.
Frequently Asked Questions
How long should a long explainer article be?
Most written explainers land between 1,200 and 2,500 words, and 2,500 to 4,000 is normal when the topic needs real background. Length follows the reader’s question, not a target: if you can answer the central question in 800 words without cutting evidence, stop at 800. One agency analysis of page-one results put the average around 1,450 words, which is a useful sanity check rather than a rule. Judge the piece by whether every paragraph earns its place, not by the counter.
Should I write a long explainer article before creating an outline?
No. Drafting a long explainer first means you build the shape while making the arguments, which is where the shape usually breaks. Spend an hour on the outline: the central reader question, the sections as questions, and a word budget for each block. Write the lede and nut graf last, once you know what the piece actually argues. Plenty of experienced writers draft straight through, but they do it on shorter pieces where the cost of rearranging is low.
How many sections should a long explainer article have?
Five to nine main sections is the working range for a 1,500 to 3,500 word explainer. Fewer than five usually means one section is carrying two arguments, and more than nine means the reader is being asked to hold too much at once. A common pattern is one section on why this matters, two or three on how it works, one on evidence, one on limits, and one on what to do. Subheadings inside those sections can be more generous, since they only interrupt one argument.
When should I add a table of contents to a long explainer?
Add one past about 1,500 words, or any time a reader would plausibly arrive mid-article from a search or a link. Below that length, a table of contents costs more space than it saves. Keep it to section-level links, skip the subheadings, and make sure the anchors are real headings rather than bold text. For a scrolling or interactive explainer, replace the table with a progress indicator instead of shipping both.
How do I keep a long explainer readable on mobile devices?
One idea per paragraph, three sentences maximum, and subheadings every 200 to 300 words so the reader can find their place by scrolling. Keep tables narrow or let them scroll sideways rather than squashing columns, and add alt text that describes what each graphic shows. Avoid long block quotes on narrow screens. Most of all, front-load: mobile readers abandon early exits more aggressively, so the lede and nut graf have to be in the first screen and a half.
Conclusion
To structure a long explainer article, start with one sentence naming the question a specific reader will be able to answer when they finish, and refuse every section that does not serve it. Lay the evidence out against that question, choose one narrative progression, and give each block a word budget instead of giving the piece a single total. Draft each section point-first, edit in three passes, and check where a tired reader would have stopped.
Most of the work happens before the first paragraph, and it is faster than rewriting a 3,000-word draft that was shaped wrong from the start.
