The Langer 1978 copy-machine study is the most-cited justification in copywriting for “always include the word because.” The original paper actually shows the placebic-reason magic disappears the moment the request stops being trivial. Here is what Langer, Blank, and Chanowitz really found, and what the copy decks omit.

Open any copywriting deck from the last twenty years and you will, sooner or later, encounter a slide that says something like this. People do not need a good reason. They just need a reason. Below the bullet, a citation to “Langer, 1978, Harvard.” Below the citation, three numbers in a column. 60% --- no reason given. 94% --- real reason given. 93% --- fake reason given. Below those numbers, a takeaway in bold: Always include the word “because” in your copy. It does not even matter what comes after.

I have seen versions of this slide in CRO trainings, in landing-page-anatomy decks, in email-marketing courses, in sales-call playbooks, in pricing-page workshops, and in at least three different “psychology of persuasion” YouTube series. The slide is so common, and the numbers so memorable, that the underlying study has become a kind of folk wisdom in the marketing world. Just say because. The brain is dumb. Compliance follows.

There is a problem with this slide, and the problem is not that the numbers are wrong. The numbers are essentially correct. Sixty percent, ninety-four percent, ninety-three percent --- those are real figures from a real study, conducted at the Graduate Center of the City University of New York, on a real photocopier, with real strangers approached by real confederates asking to cut in line. Ellen Langer, Arthur Blank, and Benzion Chanowitz published the experiment in 1978 in the Journal of Personality and Social Psychology, and the basic finding has been re-examined enough times that it is reasonable to call it replicated, at least within the conditions Langer and her co-authors tested.

The problem is that the conditions are doing almost all of the work, and the marketing slide leaves them out entirely.

Because the same paper, on the same Xerox machine, with the same confederates and the same script except for one substitution, ran a parallel set of trials with a different request size. Not five pages. Twenty pages. And in that condition --- the condition where the favor being requested actually cost the person being asked something meaningful --- the placebic-reason magic that the marketing world built a career on did not work. The compliance rate for “because I have to make copies” collapsed to within a hair of the no-reason baseline. The real reason still moved compliance; the fake one did not.

This is one of the most instructive cases in the entire replication-crisis hub. The original finding was real. The original finding was published honestly. The original authors specifically demonstrated the boundary condition that makes the finding break. And the marketing literature, almost unanimously, decided to cite the half of the study that supported the punchy slide and quietly omit the half that contradicted it. The result is two decades of copywriting advice that systematically misrepresents what the experiment showed.

This article is about what Langer 1978 actually demonstrated, what subsequent work confirmed and refined, why the elaboration-likelihood model predicts exactly the boundary Langer found, and what an honest copywriter or CRO practitioner should take away from the real evidence rather than the deck-friendly version of it.

What Langer 1978 Actually Tested

The paper is titled “The Mindlessness of Ostensibly Thoughtful Action: The Role of ‘Placebic’ Information in Interpersonal Interaction.” The full citation is Langer, E. J., Blank, A., & Chanowitz, B. (1978). The mindlessness of ostensibly thoughtful action: The role of “placebic” information in interpersonal interaction. Journal of Personality and Social Psychology, 36(6), 635—642. DOI: 10.1037/0022-3514.36.6.635.

The experiment that the marketing world cites is the first of three reported in the paper. The setup is straightforward. Confederates of the experimenters waited at one of the photocopiers in the Graduate Center library at the City University of New York. When an adult subject approached the machine and was about to use it, the confederate would step in front of them and ask to cut in line. The total sample across the experiment was 120 adults --- 68 men and 52 women --- drawn from the natural population of people who happened to be using a graduate-school library photocopier on a normal day.

The confederate’s request varied along two factors. The first factor was request size, with two levels. In the “small favor” condition, the confederate asked to copy five pages. In the “big favor” condition, the confederate asked to copy twenty pages. The second factor was request structure, with three levels.

  • Request only --- “Excuse me, I have five [twenty] pages. May I use the Xerox machine?”
  • Real information --- “Excuse me, I have five [twenty] pages. May I use the Xerox machine, because I’m in a rush?”
  • Placebic information --- “Excuse me, I have five [twenty] pages. May I use the Xerox machine, because I have to make copies?”

The placebic condition is the experimental innovation that makes this study famous. The phrase “because I have to make copies” is, on its face, content-empty --- of course the requester has to make copies, that is what one does at a photocopier. The phrase has the syntactic structure of a reason --- request followed by “because” followed by a clause --- without containing any informational substance that the listener did not already possess. The whole point of including the placebic condition was to test whether compliance was driven by the information in the reason or merely by the form of giving a reason.

The dependent measure was simple. Did the person let the confederate cut, yes or no? The cells were filled with twenty subjects each in the small-favor condition (twenty per request type, two request types --- actually with the original paper’s reporting it was slightly different per-cell ns, but the order of magnitude is correct), and a similar number in the big-favor condition. The total sample of 120 adults is small by modern standards, but the design was clean, the manipulation was simple, the dependent measure was binary, and the field-experiment setting eliminated most of the demand-characteristic concerns that haunt laboratory persuasion research.

In the small-favor condition, the results were the ones the marketing world remembers.

  • Request only --- 60% compliance.
  • Real information (“I’m in a rush”) --- 94% compliance.
  • Placebic information (“I have to make copies”) --- 93% compliance.

The placebic reason produced essentially the same compliance as the substantive reason, and both produced substantially higher compliance than asking with no reason at all. Langer’s framing of this result was that compliance was being driven not by an evaluation of the reason’s content but by mindless script-following --- the listener encountered the syntactic structure “[request] because [clause],” matched the structure to a familiar interpersonal-request script, and complied without ever processing what the “because” clause actually said. Hence the paper’s title. Ostensibly thoughtful action. The compliance behavior looked thoughtful --- it appeared to incorporate the reason given --- but the cognitive process underneath was mindless.

This is the half of the paper that survived into the marketing canon. Now the other half.

The Result Marketers Skip

The big-favor condition used the exact same three request structures, the exact same confederates, the exact same photocopier, and changed exactly one thing: the request was now to copy twenty pages instead of five.

  • Request only --- roughly 24% compliance.
  • Real information (“I’m in a rush”) --- roughly 42% compliance.
  • Placebic information (“I have to make copies”) --- roughly 24% compliance.

(The numbers above are reconstructed from Table 1 of the original paper, where the figures are reported as percentages declining the request; compliance is the complement.)

Read those three numbers in sequence and the picture changes completely. Asking with no reason produced about a quarter compliance --- a reasonable baseline for a meaningful favor being asked of a stranger in a library. Asking with a real reason --- “I’m in a rush” --- substantially increased compliance, from roughly a quarter to roughly forty percent. And asking with a placebic reason did essentially nothing. The compliance rate with “because I have to make copies” was statistically indistinguishable from the no-reason baseline. The placebic word “because” lost its magic the moment the favor stopped being trivial.

This is not a buried result. It is reported in the same table as the small-favor result, in the same paper, by the same authors. Langer and her co-authors discussed the boundary explicitly. Their interpretation was that mindless script-following operates only when the cost of complying is low enough that the listener does not bother engaging deliberate evaluation of the request. When the cost goes up, evaluation kicks in. And when evaluation kicks in, the listener notices that “because I have to make copies” does not actually contain a reason --- and the persuasive power of the syntactic structure evaporates.

The conclusion the paper itself drew was nuanced. It was not “the word because is magic.” It was “the word because is magic in low-stakes script-driven encounters, and stops being magic the instant the encounter requires the listener to think about whether to comply.”

The marketing literature largely kept the first clause and dropped the second.

Langer’s Theoretical Framework --- Mindlessness vs Mindfulness

To understand why the boundary in Langer 1978 is the actual finding, and not an inconvenient caveat to the actual finding, it helps to know what theoretical framework Langer was building in her broader research program. The copy-machine experiment was not a one-off bit of cleverness; it was a piece of evidence in a larger argument that Langer would later codify in her 1989 book Mindfulness (Addison-Wesley), and that would shape her career for the next several decades.

Langer’s argument is that human social and cognitive behavior is organized along a spectrum from mindless --- automatic, script-driven, category-bound, requiring minimal attentional engagement --- to mindful --- actively attentive to novel features of the current situation, willing to recategorize, sensitive to context. Neither end of the spectrum is universally good or bad. Mindless processing is efficient and lets people get through the day without exhausting their cognitive resources on routine interactions. Mindful processing is more accurate when the situation requires careful evaluation but is expensive enough that no one engages it constantly.

The copy-machine experiment was designed, in Langer’s framing, to demonstrate that interpersonal-compliance behavior is more often mindless than people realize. The placebic-reason result in the small-favor condition was a striking demonstration: subjects’ compliance was responsive to the syntactic form of the request rather than to the content, which meant they were not actually processing the content. That was the mindlessness claim. And the big-favor condition was the corresponding demonstration that mindfulness can be triggered by stakes --- when the cost of compliance went up, subjects shifted from script-following to evaluation, and the syntactic trick stopped working.

In other words, the boundary that marketers omit is not an inconvenient asterisk on Langer’s main result. It is part of Langer’s main result. The paper is not a paper about “the word because.” It is a paper about when humans process language mindlessly and when they shift to mindful processing --- and the placebic-reason finding and the big-favor null finding are two halves of the same theoretical demonstration. Citing one without the other does not partially capture Langer’s contribution; it inverts it. The marketing slide that says “the brain is dumb, just give it any reason” is selling Langer’s theoretical framework precisely backwards. Langer’s actual claim is that the brain is dumb in trivial encounters and stops being dumb the moment something is at stake.

This matters for any marketer trying to apply the finding, because the application Langer’s theory licenses is much more specific than the application the marketing canon assumes. The theory says: if you can keep the stakes of the request low enough that the listener does not engage evaluation, syntactic-structure cues will move compliance. The theory also says: as soon as the listener has any reason to think hard about your ask, syntactic-structure cues will lose their grip and the actual content of your justification will start to matter. Those are two different prescriptions for two different parts of a funnel.

How Petty And Cacioppo’s Elaboration-Likelihood Model Explains The Boundary

Langer’s mindlessness framework was not the only theoretical apparatus developed in this period to explain why the same persuasive technique can work in one setting and fail in another. The dominant framework in persuasion research from the mid-1980s onward was the elaboration-likelihood model, developed by Richard Petty and John Cacioppo and articulated most fully in their 1986 paper Petty, R. E., & Cacioppo, J. T. (1986). The elaboration likelihood model of persuasion. Advances in Experimental Social Psychology, 19, 123—205.

The ELM proposes that persuasive messages can influence attitudes through two qualitatively different routes. The central route involves effortful cognitive engagement with the substance of the message --- the recipient considers the arguments, weighs the evidence, evaluates the logic, and updates their attitude based on the quality of the case being made. Persuasion via the central route tends to produce attitude change that is durable, resistant to counter-argument, and predictive of later behavior. The peripheral route involves attitude change driven by cues that are not part of the substantive argument --- the attractiveness of the source, the number of arguments rather than their quality, the polish of the production, the syntactic form of the request, the presence of cues that signal expertise or authority. Persuasion via the peripheral route tends to produce attitude change that is shallower, less durable, and more easily undone by subsequent counter-information.

The crucial part of the model, for our purposes, is the prediction about when each route operates. Petty and Cacioppo argued that route selection depends on two factors: the recipient’s motivation to process the message and their ability to do so. When motivation and ability are both high --- the topic is personally relevant, the recipient has the cognitive resources to engage, no distractions are interfering --- central-route processing dominates, and persuasion succeeds or fails based on argument quality. When motivation or ability is low --- the topic feels trivial, the recipient is rushed or distracted, the encounter is brief --- peripheral-route processing dominates, and persuasion succeeds or fails based on cue presence rather than cue substance.

Now reread Langer’s copy-machine results through this lens. The small-favor condition is a low-motivation, low-elaboration setting. A stranger asks to cut five pages of copying ahead of you. The cost to you is trivial. You have no reason to engage central-route processing. You operate peripherally --- and the peripheral cue “request followed by because-clause” looks enough like a normal interpersonal-request script that you comply. Whether the because-clause actually contains a substantive reason is invisible to your processing system, because you are not engaging the system that would notice.

The big-favor condition shifts the recipient toward central-route processing. Twenty pages is a meaningful cost. You have reason to engage the question of whether this request is worth granting. Once you engage central-route processing, the peripheral cue stops mattering and the substance starts mattering --- and the substance of “because I have to make copies” is empty. So compliance falls back to the level it would be at with no reason at all.

This is exactly the prediction the elaboration-likelihood model would make, and Langer’s data fit it cleanly. The marketing application of the study --- “always include because” --- would only be valid if every persuasive context resembled the small-favor condition. Some persuasive contexts do. Most do not. And the contexts in which marketers make their hardest asks --- a pricing-page decision, a high-commitment signup, a meaningful purchase --- are exactly the contexts in which central-route processing is most likely to be engaged, which means exactly the contexts in which the placebic-reason trick will produce nothing.

The most common form of the misapplication runs something like this. A copywriter is editing a CTA from “Sign up for our newsletter” to “Sign up for our newsletter because you’ll get weekly tips.” The copywriter cites Langer 1978 as the evidentiary basis for the change. The new version has a “because” clause, the deck says “because” is magic, therefore the change must improve conversion.

The problem is that the original study does not support this inference, and not for subtle reasons. It does not support it because the “newsletter signup” decision is not a 5-page copy-machine ask. It involves the recipient evaluating whether they actually want more email from you, whether the content will be worth the inbox cost, whether your brand is one they want to be associated with, whether previous newsletter subscriptions have been useful. These are central-route considerations. The recipient, in deciding whether to sign up, is engaging the cognitive system that notices what your “because” clause says. If the clause is substantive --- “because we’ll send you the three most-shared CRO experiments each week” --- it might improve conversion. If the clause is placebic --- “because you’ll get weekly tips” --- it will not, because in central-route processing, “you’ll get weekly tips” is a tautology, the listener notices the tautology, and the syntactic-form cue does no lifting.

The marketing application that would be supported by Langer 1978 is much narrower. It would be a context in which the recipient is operating in a low-stakes, scripted, peripheral-processing mode --- say, a cookie-consent banner whose default is “accept and continue,” where the user is mid-flow toward something else and not engaging deliberate evaluation. In that kind of context, a syntactically reasoned ask might marginally outperform an unreasoned ask. But these are not the contexts that drive most marketing ROI. The places marketing copy actually has to do work --- pricing pages, signup forms, CTAs on consideration-stage content --- are central-route contexts, and they are not the contexts Langer tested.

The right takeaway from Langer 1978 is not “always include because.” It is “low-stakes asks tolerate cheap justifications; high-stakes asks need real ones.” This is a calibration heuristic, not a universal rule. It tells you that on the parts of your funnel where the user is barely paying attention, you can get away with thin reasons; on the parts where the user is evaluating you, you need substance. That is a much more useful framework than the slide-deck version, and it has the additional virtue of being what the study actually shows.

What Folkes 1985 Added --- And Why It Matters

Seven years after Langer’s original paper, Valerie Folkes published a follow-up that complicated the picture further: Folkes, V. S. (1985). Mindlessness or mindfulness: A partial replication and extension of Langer, Blank, and Chanowitz. Journal of Personality and Social Psychology, 48(3), 600—604. DOI: 10.1037/0022-3514.48.3.600.

Folkes ran the copy-machine paradigm again, with one variation. Instead of varying only whether the reason was placebic or substantive, she varied whether the reason invoked something the requester could control versus something they could not. “Because I don’t want to wait” is a reason, but it invokes the requester’s controllable preferences. “Because I feel really sick” is also a reason, and it invokes something outside the requester’s control.

Her finding was that compliance was substantially higher when the reason was uncontrollable than when it was controllable, even though both had the syntactic structure of a reason. This is not what a pure mindlessness account would predict --- if the only thing that mattered was the syntactic presence of “because,” controllable and uncontrollable reasons should produce equivalent compliance. They did not. Listeners were processing the content of the reason enough to distinguish between “you have a preference” and “you have a need,” and weighting their compliance accordingly.

Folkes ran additional studies and concluded, in her own words, that “although everyday social interactions may often be less than completely mindful, they do not appear to be as mindless (automatic) as might have been surmised.” Her studies did not replicate the original Langer placebic-equals-substantive finding under all conditions; in particular, when she used a controllable reason as her “substantive” condition, that reason underperformed the placebic one because listeners actively disliked having a controllable preference imposed on them.

The Folkes result is important for two reasons. First, it confirms what Langer’s own big-favor condition had already suggested: there is more cognitive processing of compliance requests than the pure mindlessness account would have you believe, and content matters more than the marketing version of the study admits. Second, it shows that when reasons do matter --- when central-route processing is engaged --- the type of reason can move compliance in either direction. Reasons that invoke uncontrollable needs help compliance; reasons that invoke controllable preferences can actually hurt it. This is information no honest copywriter should ignore, and it is information that the standard “Langer proved you just need a reason” deck systematically suppresses.

The honest synthesis of Langer 1978 plus Folkes 1985 looks something like this. In trivial encounters where the listener is operating mindlessly, the presence of a “because” clause moves compliance regardless of content. In meaningful encounters where the listener is engaging substantive evaluation, the content of the reason matters --- and the wrong kind of reason can hurt rather than help. Neither of these claims supports the slide-deck version that has dominated marketing folklore for forty years.

What’s Honest To Say About Reasons In Persuasion Now

Strip away the marketing oversimplification and the academic-vs-popular gap, and the honest summary of the evidence on reasons in persuasion looks like this.

The original Langer 1978 finding is real, in the conditions it was tested in. In low-stakes scripted interpersonal-compliance encounters, providing a request with a syntactic-reason structure increases compliance over providing a request alone, and the increase is largely insensitive to whether the reason has informational content. This part of the finding has held up reasonably well in subsequent work within the same kind of setting.

The boundary condition that Langer herself reported is also real, and is the more important half of the result. When the cost of compliance increases enough that listeners shift from peripheral to central processing, the syntactic-form effect collapses. Empty reasons stop working. Substantive reasons continue to work, but only because they are substantive --- not because they contain the word “because.”

The Folkes 1985 extension complicates the simple “any reason works” story further. When listeners engage content processing, they not only distinguish empty reasons from substantive ones, they also distinguish controllable reasons (which can hurt compliance) from uncontrollable ones (which help). The type of reason matters in central-route conditions in ways that the original study did not investigate.

The elaboration-likelihood model explains both Langer’s boundary and Folkes’ extension cleanly. Peripheral-route persuasion is driven by form cues including the syntactic presence of reasons; central-route persuasion is driven by content quality. Whether a given persuasive context is peripheral or central depends on the recipient’s motivation and ability to engage substantive processing, both of which are higher when the stakes are higher.

There is, to my knowledge, no rigorous A/B testing literature that confirms the popular “always include because” prescription in any commercially meaningful conversion setting. There are many copywriting case studies and blog posts that claim it works, but a case study with no control, no statistical power calculation, no preregistration, and no replication is approximately the same evidence weight as a personal anecdote. The actual experimental literature, when it has examined the question, has tended to find that the effect of reason-giving in marketing copy is heavily conditional on context, audience, and stakes --- exactly what the Langer-plus-Folkes-plus-ELM synthesis would predict.

The marketing claim “always include the word because in your copy because it doesn’t matter what comes after” is, in the strictest possible interpretation, not supported by the evidence base it cites. The cited evidence base actually supports the substantially more nuanced claim “in low-stakes peripheral-processing contexts, the syntactic form of giving a reason can marginally move compliance; in higher-stakes central-processing contexts, the content of the reason matters more than its form.” The fact that the second sentence is harder to put on a slide is, presumably, why the first one won.

What This Means For Your Copy And Conversion Programs

Here is the calibration framework that actually maps onto the evidence.

Identify where in your funnel the user is operating mindlessly versus mindfully. Top-of-funnel content consumption, scrolling through an email, dismissing a cookie banner, deciding whether to click a CTA that has not yet asked them to commit to anything significant --- these are peripheral-processing contexts. The user is in a low-elaboration mode. Syntactic cues, including the presence of reasons that may or may not be substantive, are likely to do some lifting. This is where Langer’s small-favor finding plausibly applies.

Identify where in your funnel the user is being asked to commit to something meaningful. A pricing-page decision, a signup form with real fields, an enterprise contact request, a checkout step, a paid-plan upgrade --- these are central-processing contexts. The user is engaging substantive evaluation. The content of your reasons matters enormously; the syntactic form of giving them matters very little. This is where Langer’s big-favor finding applies, and where the placebic-reason trick will produce nothing.

Treat “because” as a syntactic structure, not as a magic word. If your reason is substantive, using “because” to introduce it may marginally improve readability and signal that you are about to give a justification. If your reason is empty, dressing it up in a “because” structure will not save it in any context where the user is paying attention.

A/B test the content of your reasons, not just their presence. The interesting question for any high-stakes ask is not “should I include a reason.” Of course you should. The interesting question is which reason produces the highest lift, and that requires actual experimentation with substantive variation in the content. Testing “Sign up” against “Sign up because we’ll send you tips” is the wrong test. Testing “Sign up because we’ll send you the three most-cited experiments from the last week” against “Sign up because 12,000 PMs already use this for their weekly planning” against “Sign up because we’ll cancel your subscription within ten seconds if you do not find it useful” is the right test.

Be especially careful with controllable-preference framings. Folkes 1985 suggests that reasons invoking the requester’s controllable preferences can hurt compliance relative to reasons invoking uncontrollable needs. In marketing translation: “we’d love it if you’d sign up because it would help our metrics” is asking the user to do something for your controllable preferences, and will probably underperform “we’d love it if you’d sign up because we think you’ll find a specific thing useful.” The mechanism is the same as in Folkes’ library subjects: when listeners process the content of a reason, they evaluate whether the requester is asking them to subsidize a controllable choice or accommodate an uncontrollable need, and they react differently.

Update your team’s mental model of the Langer study. If your copywriting onboarding deck currently cites Langer 1978 as evidence that “always include because” is a universal rule, the deck is misrepresenting the cited study. The honest version of the slide includes the big-favor result, the Folkes 1985 complication, and the ELM framework that explains both. The honest version may be less viscerally satisfying than the magic-word version, but it is the one that will actually help your team build copy that works.

The deeper lesson here is the same one that runs through this hub. Popular psychology overgeneralizes from narrow experiments to universal prescriptions, and the original studies are often more honest about their conditions than the secondary literature that cites them. Langer 1978 reported its boundary condition openly in the same paper as its headline finding. The marketing canon, over forty years, kept the headline and dropped the boundary. That is a failure of citation discipline, not a failure of the original science --- and the corrective is for marketers to read the actual paper before quoting the actual numbers.

Sources

  • Langer, E. J., Blank, A., & Chanowitz, B. (1978). The mindlessness of ostensibly thoughtful action: The role of “placebic” information in interpersonal interaction. Journal of Personality and Social Psychology, 36(6), 635—642. DOI: 10.1037/0022-3514.36.6.635
  • Folkes, V. S. (1985). Mindlessness or mindfulness: A partial replication and extension of Langer, Blank, and Chanowitz. Journal of Personality and Social Psychology, 48(3), 600—604. DOI: 10.1037/0022-3514.48.3.600
  • Langer, E. J. (1989). Mindfulness. Reading, MA: Addison-Wesley. (Theoretical context for the mindless/mindful processing distinction.)
  • Petty, R. E., & Cacioppo, J. T. (1986). The elaboration likelihood model of persuasion. Advances in Experimental Social Psychology, 19, 123—205. (The two-route persuasion framework that predicts Langer’s boundary condition.)
  • Liberman, M. “Generalization and truth.” Language Log (commentary on Langer 1978 results table). URL: https://languagelog.ldc.upenn.edu/nll/?p=1396

FAQ

Should I stop using the word “because” in my copy?

No. The word itself is fine and often improves readability when it introduces a substantive justification. The thing to stop doing is treating “because” as a magic syntactic trigger that improves compliance regardless of what follows it. In any context where your reader is actively evaluating whether to comply --- pricing pages, signup forms, paid upgrades --- the content of the reason is doing the work, not the word “because.” Use “because” when you have a substantive reason to introduce; do not use it as a fig leaf to dress up an empty claim.

Does the “because” effect work in email subject lines?

There is no rigorous evidence that adding “because” to a subject line improves open rates in the way the small-favor copy-machine condition suggests. Subject lines are a partially central-route context --- the recipient is deciding whether the email is worth opening based on the substance of what it promises --- and the placebic-reason trick should produce essentially nothing there. If you want to test reasons in subject lines, test substantive reasons against each other rather than testing “no reason” against “any reason at all.”

What about pricing-page justifications? Should those use “because” framings?

Pricing pages are a high-stakes central-processing context --- exactly the condition where Langer’s big-favor result shows the placebic trick does not work. If you want to justify your pricing, the justification needs substance: what specifically the user gets, why the price is structured the way it is, what the comparison set looks like. A vacuous “this price is set this way because we believe in great value” will be transparent to a reader who is actively evaluating whether to pay you. Substantive reasons help on pricing pages; empty syntactic markers do not.

What about CTAs? “Click here because [X]” --- does the structure help?

It depends entirely on how much the click commits the user to. A click on a top-of-funnel “Learn more” with no commitment is peripheral-processing-friendly, and a thin reason might marginally improve click-through. A click on “Start your free trial” requires the user to think about whether they want to start a trial, which is central-processing-friendly, and a thin reason will not help. The general rule: the higher the commitment behind the CTA, the more the content of any justification matters and the less the syntactic form of “because” matters.

Has anyone done a modern A/B test of Langer’s exact finding in a commercial setting?

Not that I am aware of in a publicly reported, rigorously controlled, and adequately powered form. There are many case-study posts claiming “because” lifted conversion by some headline number, but most have no control, no preregistration, and no replication --- they are not equivalent evidence to the original Langer study, let alone to a meta-analysis. If you run such a test yourself in your own funnel, you will probably find that the effect is highly conditional on which page you tested it on, which segment was exposed, and what the rest of the copy was doing. That is exactly the prediction the ELM framework would make.

Is Langer’s mindfulness work in general trustworthy?

The 1978 copy-machine experiment is one piece of a much larger research program. Some of Langer’s other findings have been more controversial --- the “counterclockwise study” on aging, for instance, has been the subject of substantial methodological critique and has not been cleanly replicated. The mindfulness framework as a general organizing concept has been productive in social and cognitive psychology, but specific empirical claims within it vary substantially in how well they have held up. The copy-machine result, in the bounded form I described above, is one of the more robust pieces of the program.

What’s the single most important takeaway for a marketer reading this?

The Langer study does not justify “always include a reason.” It justifies “in low-stakes peripheral-processing contexts, reasons help even when they are thin; in high-stakes central-processing contexts, only substantive reasons help.” That is a calibration framework, not a universal rule. Apply it to your funnel by mapping each stage to where on the peripheral-central spectrum your user is likely operating, and adjusting your copy accordingly. The decks that say “the brain is dumb, just say because” are misrepresenting the underlying science in a way that will systematically lead you to under-invest in the substantive parts of your copy where it matters most.

Why do so many copywriting authorities repeat the simplified version?

The simplified version makes a great slide. It contains a memorable number (93%), a counterintuitive twist (“the reason doesn’t even have to make sense”), an authority anchor (Harvard, although the actual study was done at CUNY), and a punchy prescription (always use “because”). The honest version requires explaining boundary conditions, replication caveats, theoretical frameworks, and stakes-based calibration --- which is harder to compress, less viscerally satisfying, and worse on social media. The selection pressure on marketing content favors the simplified version regardless of whether it accurately represents the underlying evidence. That selection pressure is part of what this hub is trying to push back against.

replication-crisis persuasion copywriting langer-1978 evidence-evaluation

Free Tool

Built for Experimentation Teams

GrowthLayer is the experimentation platform I built for CRO teams --- test management, AI-powered insights, and pattern recognition across your entire program.

Explore GrowthLayer → (opens in new tab)

· Start Free →

Share this article

LinkedIn (opens in new tab) X / Twitter (opens in new tab)

Copy link

Go deeper

Methodology The PRISM Method Case Studies $30M+ in Results Work Together Services & Mentoring

Experimentation and growth leader. Builds AI-powered tools, runs conversion programs, and writes about economics, behavioral science, and shipping faster.

About LinkedIn Newsletter

← Previous

The Default Effect: The Behavioral-Economics Finding That Actually Holds Up

Next →

Foot-In-The-Door: The Persuasion Tactic That Works (Just Less Than Marketers Think)

replication-crisis persuasion copywriting langer-1978 evidence-evaluation
Share this article
LinkedIn (opens in new tab) X / Twitter (opens in new tab)
Atticus Li

Experimentation and growth leader. CXL-certified CRO practitioner, Mindworx-certified behavioral economist (1 of ~1,000 worldwide). 200+ A/B tests across energy, SaaS, fintech, e-commerce, and marketplace verticals.

About LinkedIn Newsletter