A title is the shortest part of the work and the part most readers will ever see. It has to describe the piece, promise something specific, and set the reader's expectations for the first paragraph. Most titles fail because the first one written is the one that ships.
- A title must name the subject, the payoff, and the kind of piece it introduces.
- Never let a working title graduate to publication without being retested.
- Check that the draft actually delivers whatever the title promised.
- Write twenty candidates before you allow yourself to choose one.
- Move the caveats into the deck so the title can stay specific and short.
A title makes three promises
A title says what the piece is about, what the reader gets from it, and what kind of thing it is — the same three questions a writing brief settles before you draft. Miss any of the three and the reader has to open the piece to find out. Most of them will not open it.
That third promise is the forgotten one. A reader needs to know whether they are being handed an argument, a story, a procedure, or a warning. The expectation shapes how they read the opening lines, and a mismatch feels like a broken piece rather than a mislabelled one.
Working titles and published titles are different tools
A working title exists for the writer. The published one belongs to the publishing stage. Its job is to hold the piece steady while you draft, so it should be blunt, plain, and true. Something like the Tuesday review problem is an excellent working title and a poor published one.
A published title exists for a stranger. It has to survive with no context, sitting in a list beside other titles, on a small screen, six months later. Writers get caught out because a working title stops looking provisional somewhere around month four.
A title that repels the wrong reader is doing half its job.
Test the title against the piece it labels
Read the title, then read the final page. If the piece never delivers what the title promised, retitling is only one of your two options. Sometimes the title is right and the draft has fallen short of it, and you would rather know that before publication.
The sharper test uses someone else. Show them the title alone and ask what they expect to find. Then let them read it. Wherever expectation and delivery diverge, you have found either a title problem or a structure problem, and the difference is usually obvious once named.
Why curiosity gaps fail readers
A curiosity gap withholds the thing the reader wants so that they have to click for it. It works once. The bill arrives on the page, when the reader discovers that the answer was a single sentence and the rest was delay dressed as suspense.
Withholding also destroys the title's filtering job. A title that names its subject sends the wrong readers away, which is a feature and not a loss. A title built on what nobody tells you attracts everyone equally, so the readers who needed the piece are outnumbered.
Curiosity is fine when the answer is real and the piece hands it over early. Here is the test. Ask whether you would be comfortable putting the answer in the subheading. If that would ruin the title, the title was living on the delay.
Generate more options than you need
The first title is rarely the best one, and it usually ships anyway, because writing it feels like finishing. Force a list instead. Aim for twenty candidates in one sitting, and let most of them be bad on purpose. Bad options mark the edges of the range.
Vary the angle deliberately: name the event, the cost, the reader, the object, the surprise. Strut's title-ideas operation works from the draft itself and returns grouped options with a one-line rationale for each, plus the one it would pick. Use it to widen the list, then choose by hand.
Read the shortlist aloud inside a sentence, as though recommending it to somebody. Titles that sound absurd when spoken are usually straining for effect. The survivor is often plainer than the one you were hoping for, and it will still make sense next year.
Let the deck carry the qualifications
A title should not have to hold every caveat. That is what a deck, standfirst, or subtitle is for. Put the event in the title and the price in the deck. The pair can then be precise without either one turning into a paragraph.
The same split works inside a book. A short chapter title beside a one-line summary in the contents gives the reader the shape of the argument, and it saves you from writing chapter headings that try to explain themselves in eleven words.
Choose, then stop
Pick the most specific title that is still true. Specificity is what makes a title findable, quotable, and hard to confuse with the four other pieces published that week on the same broad subject. Vagueness is not neutrality; it is a lost reader.
Then stop. A title can be tinkered with indefinitely, and past some point you are only rearranging words rather than sharpening a promise. Sleep on the shortlist once, choose in the morning, and hand the piece over to the reader it was written for.
The same 4,200 words, titled twice
As originally published, on a piece about a distribution centre that stopped producing demand forecasts: "The Road Ahead: Reimagining Supply Chain Excellence in a Volatile Age" Deck: "In a fast-moving landscape, forecasting matters more than ever. We share our journey and our key learnings." The piece underneath is narrow and useful: six weeks in one warehouse, the three forecasts the team stopped producing, the order-data rule they used instead, the two stockouts that followed, and the smaller forecast they eventually restored. None of that is visible.
The retitled version, same text, one changed paragraph at the top: "We ran a warehouse for six weeks with no demand forecast" Deck: "What we used instead, the two stockouts it cost us, and the smaller forecast we went back to." Other candidates from the list of twenty: "The three forecasts we stopped producing"; "Six weeks, no forecast, two stockouts". The chosen title states the event in plain words. The deck states the price and the ending, so nobody arrives expecting a manifesto.
The first title promised a category and delivered a case study, disappointing both audiences. The second promises one event, and the deck absorbs the qualifications a title cannot carry. The ending is given away deliberately, so a reader who only wanted the answer gets it honestly.
Common mistakes
| The mistake | Why it costs you, and what to do instead |
|---|---|
| Publishing the working title | It stops looking provisional after a few months, and by then you are too close to hear it. Write the published title as a separate task, after the draft is finished, and compare it against at least ten alternatives. |
| Promising a category instead of a piece | Titles about the future of a whole field attract people expecting a survey and repel the ones who wanted your specific case. Name the thing you actually did or found, at the size you actually did it. |
| Building the title on a gap you never fill | Withheld answers spend the reader's goodwill before the first paragraph. If the payoff cannot appear in the deck without wrecking the title, the title is selling suspense rather than the piece. |
| Choosing the first option that is not embarrassing | Relief is not a selection criterion, and it arrives early. Keep writing candidates past the point of comfort, then leave the list overnight. The one that still reads well in the morning is usually a different one. |
Your checklist
- Write a blunt working title before drafting, and mark it as provisional.
- Draft twenty published candidates in one sitting once the piece is finished.
- Cover at least five angles: event, cost, reader, object, and surprise.
- Read the shortlist aloud and cut anything that sounds strained.
- Move every caveat out of the title and into the deck.
- Confirm the piece delivers the promise on its first page.
Key terms
- Deck
- The short line of text sitting under a title, also called a standfirst. It carries the qualifications, the scope, and the price of the promise the title makes.
- Working title
- A provisional title used during drafting to keep the writer aimed at one subject. It is written for the writer and is not expected to survive publication.
- Curiosity gap
- A title built by withholding the information the reader wants, so that opening the piece becomes the only way to get it. It converts attention once and spends trust each time.
Working on a draft with AI right now?
Frequently asked questions
Should a title contain search terms?
+
Usually it already does, because the phrase people search is often the plainest description of the subject. Write the honest, specific title first, then check that the words a reader would type appear in it. Do not bolt on terms that make the sentence read like a listing.
How long can a title be?
+
Long enough to be specific, short enough to scan in a list. The practical constraint is that the distinguishing words must arrive early, because titles get truncated in feeds, tabs, and search results. Front-load the subject and let any qualifier trail.
Should I title before or after writing?
+
Both, for different reasons. A working title before drafting keeps the piece from wandering. A published title afterwards can describe what the piece became, which is rarely what you set out to write. Confusing the two is the most common titling error.
Do the title, heading, and web address need to match?
+
Keep the visible title and the main heading identical, or readers assume they clicked the wrong link. The web address can stay stable even after a retitle. Changing addresses breaks links other people already made, and a stale address costs nobody anything.
Can I retitle a piece after publication?
+
Yes, and it is often worth doing once you can see how people describe the piece back to you. Keep the address, update the heading and any internal references, and leave a short note if the change alters what the piece appears to promise.
How do I test a title with no audience?
+
Show the title alone to one colleague and ask what they expect to find. Then place each candidate in a list of five real titles from your field and see which one you would open. None of that is a study, and all of it beats guessing.