There is a familiar piece of copywriting advice: if you want someone to keep reading, ask a question. Questions appear to create unfinished business, so once something has been asked, the reader should want the answer.
It is why titles across marketing, journalism and social media so often begin with “Why”, “Could”, “Should” or “Is”. A statement closes something down, while a question seems to open it up. The question creates a gap and curiosity is expected to pull the reader across.
Research by David Fang and S. Christian Wheeler, published in the Journal of Consumer Psychology in July 2026, suggests the sequence can work differently. Across four studies, titles framed as questions received less engagement than titles framed as statements. Their explanation is not that readers dislike questions. It is that a question can make the content behind it appear less informative.
The distinction changes the problem because, before deciding whether an unanswered question is intriguing, a reader may first decide whether the answer is likely to be worth their time.
The question headline makes intuitive sense
The case for questions begins with a sound idea about curiosity. When people become aware of a gap between what they know and what they want to know, resolving that gap can become rewarding in its own right. A title such as “Is your marketing strategy working?” names something unresolved and invites the reader to complete it.
That logic helps explain why question headlines are so common. They address the reader directly, allow a subject to be introduced without giving away the conclusion and can make a piece of content feel like an inquiry rather than an announcement. For marketers, they also offer an easy way to turn almost any topic into an apparent reason to click.
Yet a title is not encountered in isolation. It usually appears beside other titles, each competing for a finite amount of time. In that setting, readers are not only deciding what makes them wonder. They are estimating what each piece will give them in return.
Fang and Wheeler focus on this prior judgement. A question may make an answer unknown while also weakening the title's informational value. The curiosity gap then has two effects working in opposite directions: it creates something unresolved but gives the reader less evidence that resolving it will be useful.
Four studies examine four kinds of engagement
The breadth of the research matters because “engagement” does not mean the same thing in every setting. The authors did not rely on one group choosing between hypothetical headlines. Their programme moves from naturally occurring content to live headline tests and then to a controlled experiment designed to investigate the mechanism.
Study One A analysed 53,030 Reddit posts from r/WallStreetBets. Titles containing question marks were associated with fewer comments and lower Reddit scores. Study One B examined 3,078,791 academic articles and found that papers with question titles were associated with fewer citations. These large observational analyses show that the pattern appears in two very different bodies of writing, but they cannot establish that the question framing alone caused the difference. The authors' models can account for measured features, not every reason one post or paper might attract attention.
Study Two used 22,743 online news headline A/B experiments from the Upworthy Research Archive. Visitors in these experiments were shown alternative headlines for the same article and engagement was measured as click-through rate, calculated from clicks divided by impressions. Question-framed variants produced lower click-through rates across those experiments. Because exposure to headline variants occurred within real publishing experiments, this evidence comes closer to the decision marketers care about: whether changing the title changes the likelihood of a click.
Study Three was a preregistered laboratory experiment with 400 participants. Each participant saw one of four news titles, presented either as a statement or as a question, then rated their interest in reading the article and their likelihood of doing so. The researchers also measured expected informativeness, curiosity and expected reading effort. This study traded scale for control, allowing the title frame to change while the underlying words and topic remained the same.
Taken together, the studies do different jobs. The archives establish that the pattern exists in naturally occurring content. The news experiments show it in real choices between headline variants. The laboratory study isolates question framing and tests why it might matter. None is decisive alone, but the convergence is harder to dismiss as a quirk of one platform or one measure.
The question mark is not the mechanism
The paper's most useful contribution is not the result that statements did better, but the proposed explanation for the difference.
The authors' green tea example makes the mechanism concrete. “Green Tea Has Health Benefits” already communicates a proposition. “Does Green Tea Have Health Benefits?” identifies the subject but withholds the proposition. Both titles concern the same issue and either could lead to an informative article, yet the statement offers evidence of what the reader will learn before asking for attention.
In Study Three, participants shown question-framed titles reported lower expected informativeness as well as lower intended engagement. The statistical mediation analysis was consistent with perceived informativeness carrying the effect of title framing on reading interest. By contrast, question framing did not produce a reliable increase in the study's measure of curiosity, which asked whether the headline created a gap between what participants knew and wanted to know.
That result deserves careful wording because a mediation analysis supports the authors' account without proving that every reader consciously weighs informativeness before clicking. It does show why “people do not like questions” is an inadequate summary. The format altered what participants expected the article to provide.
Nor does the finding mean that punctuation itself has a mysterious effect. In the controlled study, adding a question mark turned a claim into something uncertain. The relevant shift was communicative: the statement supplied information, while the question asked the article to earn a click before supplying it.
Curiosity needs a reason
Marketers often describe curiosity as though uncertainty automatically creates it. The research suggests a more demanding relationship. An unanswered question can become a reason to engage, but only if the reader values the possible answer.
This is partly an interpretation of the evidence rather than a separate result tested across every study. Still, it fits the mechanism Fang and Wheeler identify. In a crowded feed or search result, readers are choosing between competing promises of knowledge. A vague question may open an information gap while revealing so little about the likely return that the gap is easy to ignore.
Statements can perform a subtler role because giving away one useful idea does not necessarily remove the reason to continue. It can demonstrate that the writer has something definite to say, help the reader judge relevance and create a more precise gap around the explanation or evidence. “Question titles reduce engagement” resolves the broad issue but leaves the method, mechanism and boundary conditions worth discovering.
This does not invalidate the wider psychology of curiosity. It adds a condition that headline advice often overlooks. Uncertainty has to attach to something the reader already considers valuable. Withholding the central proposition may sometimes stimulate interest, but it can also remove the signal that makes further attention feel worthwhile.
A title is a promise of information
The practical implication is not a ban on question headlines. The paper studies titles for written content and reports an average pattern across particular datasets and experiments. It does not establish that every statement will outperform every question or that the same result transfers unchanged to email subject lines, advertisements and other formats.
It does, however, give marketing teams a better question to ask about titles. Alongside whether a title stops the scroll or creates intrigue, teams can examine how clearly it communicates what someone will know after reading. That is especially relevant to thought leadership, research reports, editorial marketing, LinkedIn posts and educational content, where the value of the material depends on a credible promise of insight.
This also has implications for testing because a question and a statement should not be treated merely as stylistic alternatives. They can establish different expectations about how much information follows. Testing headline variants should therefore measure the behaviour that matters, while preserving the topic and proposition closely enough to understand what changed.
Internal reactions to a title are a poor substitute. A question can sound clever in a review meeting because everyone present already knows the answer and the value of the content. A reader encountering the title cold has neither advantage. What feels intriguing to the team may look noncommittal to the audience.
Marketing has spent years learning how to open curiosity gaps. Fang and Wheeler's research suggests the more useful skill may be judging how much of the gap a title needs to close. The strongest title may not leave the reader with the biggest unanswered question. It may give away enough to show why the rest is worth knowing.
What teams need to know
What did the researchers find?
Across two archival studies, real-world news headline experiments and a controlled laboratory study, question-framed titles were associated with or produced lower engagement than statement-framed titles.
How was engagement measured?
The measures differed by study: Reddit comments and scores, academic citations, news click-through rate and participants' reported interest and likelihood of reading an article.
Why might question titles reduce engagement?
The laboratory evidence suggests that readers expect question-framed titles to be less informative. A question can create uncertainty while providing less evidence that its answer will be useful.
Does the research mean marketers should never use a question title?
The evidence supports an average effect, not a universal prohibition. Topic, audience and wording still matter, so a specific question may outperform a weak statement in a particular context.
Does the finding apply to email subject lines or advertisements?
The studies concern titles for written content, so applying the result to other formats is an implication to test rather than an established finding.