Intro
A pill shape with text inside it. In Product Design, this shape makes a single, binding promise: click me, and something happens. In marketing, the same shape does something else entirely. It tells you what to do next, even when "next" is somewhere else, on another screen, in another moment.
Both uses are intentional. Both are defensible. And the tension between them is worth taking seriously, because it reveals something that most CTA discourse skips over: context determines whether a visual element is an interface component or a psychological cue. Treating one as if it were the other is where the real mistakes happen.
The Disagreement
This started as a workplace debate. While working in the crypto and web3 space, I found myself in a direct disagreement with my boss about the use of CTAs across our output. His position was clear: a CTA must be functional. If it cannot be clicked, it should not look like it can be. Anything else is a failure of user experience.
My position was different. In motion graphics and static marketing assets, the pill shape serves a distinct purpose. It is not pretending to be interactive. It is priming the viewer for an interaction that exists elsewhere: on the landing page, in the app, at the end of the funnel. The button shape, in this context, becomes a visual signpost rather than a functional component.
We did not agree. We moved forward with his call. But the question stayed with me, because neither of us had the research to back the argument properly.
So I went looking for it.
Affordance, Signifier, and the Space Between
The theoretical ground for this debate was laid decades ago, and the terminology matters.
James J. Gibson introduced the concept of "affordance" in 1977 to describe the action possibilities an environment offers to an actor. A flat surface affords sitting. A handle affords grasping. These properties exist regardless of whether the actor perceives them.
Don Norman, co-founder of the Nielsen Norman Group, adapted this concept for design in 1988 with The Design of Everyday Things. Norman drew a critical distinction between real affordances and perceived affordances. In digital interfaces, real affordances are almost irrelevant: every pixel on a screen can technically be clicked. What matters is perceived affordance: does the element look interactive?
Norman later refined this further, coining the term "signifier" to describe any perceivable cue that communicates appropriate behaviour to a person. As he put it: affordances define what actions are possible, signifiers communicate where the action should take place.
William Gaver, writing in CHI '91, added another layer. He categorised affordances into three types: perceptible (visible and functional), hidden (functional but not visible), and false (visible but not functional). A false affordance is an element that looks interactive but does nothing when engaged. Gaver's framework treats this as a design failure, and in the context of product interfaces, it is.
But here is where the conversation gets interesting.
The Product Design Position
In product design, the case against non-functional CTAs is airtight. The Nielsen Norman Group's research on button states makes this explicit: an enabled button must signal through high contrast and visual prominence that it is ready to perform a specific action. A disabled button must visually communicate unavailability. Every state carries meaning, and misrepresenting that meaning erodes user trust.
NN/g's "A Link is a Promise" principle reinforces this further. A clickable element sets an expectation. When that expectation is broken, the user does not simply feel confused. They feel misled. This is not a subtle distinction. Trust in an interface is cumulative, and false affordances spend it faster than almost any other error.
The flat design movement, which stripped away the shadows, gradients, and dimensional cues that historically communicated clickability, demonstrated what happens when signifiers are weakened even unintentionally. NN/g's research on clickability found that removing 3D effects was one of the most damaging trends for perceived affordance: users could no longer reliably distinguish interactive elements from decorative ones.
Within a product, then, a button that does not function as a button is not a creative choice. It is a system failure.
The Marketing Position
Marketing and motion design operate in a fundamentally different context, and this is the point most product designers miss when they apply interface rules to non-interface media.
A CTA in a social ad, a motion graphic, or a static banner is not an interface component. The viewer cannot click it. The viewer knows they cannot click it. The medium itself makes this clear: a video playing on a screen, a static image in a feed, a billboard on a wall. There is no ambiguity about interactivity because the context removes it entirely.
What the button shape does in this context is something else. It functions as a visual primer: a recognisable cue that activates the viewer's existing association between that shape and a specific action.
The theoretical basis for this sits in cognitive psychology. Priming, as established by Bargh et al. (1996) and the broader body of research into automatic cognition, describes a process where exposure to one stimulus influences the response to a subsequent stimulus, often without conscious awareness. Schmitt's 1994 study on contextual priming of visual information in advertisements demonstrated that visual components in ads influence perception and subsequent behaviour, and that participants were not consciously aware of the influence. The mechanism is well documented. What has not been formally studied, as of writing, is the specific application of this mechanism to non-interactive CTA shapes across media formats.
The closest articulation from industry comes from KORTX's research into ad format strategy, which notes that Connected TV ads cannot offer a clickable CTA. In these formats, CTAs function as memorable visual cues: directional signals that the viewer carries with them to a second device, where the actual interaction occurs. The button shape, in this context, is not pretending to be functional. It is encoding intent.
What the Market Actually Does
To test whether this is an outlier practice or a widespread convention, I conducted my own observational research across social media platforms, collecting examples of CTA usage in non-interactive media from established brands and advertisers.
The findings are consistent and difficult to ignore.
Across static ads, motion graphics, sponsored content, and video assets, the pill-shaped CTA appears routinely in contexts where no click is possible. These are not small brands cutting corners. These are companies with dedicated product and marketing teams who understand the distinction between an interface component and an advertising asset, and who have made a deliberate choice to use the button shape as a visual signpost.
The pattern holds across industries: fintech, SaaS, e-commerce, entertainment, gaming. The CTA shape appears in Instagram carousels, YouTube pre-rolls, TikTok ads, Twitter/X promoted content, and static display banners. In every case, the medium makes interactivity impossible. In every case, the button shape is present anyway.
This is not accidental. It is convention. And it exists because the practice works: the CTA shape primes recognition, anchors the viewer's attention on the intended action, and creates visual continuity between the ad and the landing page that follows. The industry term for this continuity is "conversion scent," the principle of keeping written and visual cues consistent across the entire consumer journey so that the transition from ad to destination feels coherent rather than dislocating.
The product design objection, that a non-functional button is always a false affordance, does not account for the fact that these elements exist outside any interactive system. No user is attempting to click a YouTube pre-roll. No viewer mistakes an Instagram carousel image for a live interface. The medium has already communicated the rules of engagement. The button shape operates within those rules, not against them.
What the Research Actually Supports
The honest answer is that both positions are correct, but within their own boundaries, and the evidence supporting each comes from different places.
Gaver's framework of false affordances applies directly to interactive systems, where a visual cue that suggests clickability without delivering it constitutes a genuine failure. Norman's signifier theory supports this: in a product, every visual element must communicate its actual state truthfully. The research here is robust, well cited, and unambiguous.
The marketing position rests on a different foundation. The priming mechanism is empirically established in cognitive psychology. Its application to visual advertising is supported by Schmitt (1994) and the wider body of contextual priming research. But the specific question of whether a non-interactive CTA shape increases conversion on a subsequent interactive CTA has not been isolated in a controlled study. What exists instead is overwhelming market evidence: the practice is standard across virtually every major advertiser operating in digital media. The observational research documented in this article reflects what any audit of social media advertising will confirm.
The stronger framework for understanding CTAs in marketing is not affordance theory at all. It is priming theory, supported by observable industry convention. The button shape in a marketing asset is the prime. The actual interactive button on the landing page is the target. The effectiveness of the prime depends on its recognisability, not its functionality.
Where This Leaves the Practitioner
The answer is not one rule. It is two, applied correctly.
In a product interface, a CTA must be functional. Its visual treatment must accurately represent its state. A button that cannot be clicked must not look like one that can. This is not opinion. This is what the research supports, and violating it costs user trust.
In marketing and motion design, a CTA shape can legitimately exist as a visual cue that primes the viewer for a future interaction. Its purpose is directional, not functional. Its effectiveness depends on context: the viewer must understand, through the medium itself, that they are not being asked to click. They are being told where to click, later.
The mistake is treating these as the same discipline. Product design and marketing design share a visual language but serve different cognitive purposes. A non-functional button in an app is a bug. A non-functional button in a video is a signpost. Both are valid. Neither invalidates the other.
The discipline is knowing which context you are designing for and refusing to let the rules of one override the logic of the other.
Key Takeaways
A CTA in a product interface must always be functional; its visual treatment is a promise of interactivity, and breaking that promise erodes user trust.
A CTA shape in non-interactive media (motion graphics, static ads, video) serves as a visual primer, not a false affordance, because the medium itself removes any ambiguity about clickability.
Don Norman's distinction between affordances and signifiers is essential vocabulary for this debate: affordances define what is possible, signifiers communicate where action occurs.
William Gaver's concept of "false affordance" applies to interactive systems, not to marketing assets where no interactivity is implied by the medium.
The real error is applying the rules of one discipline to the other. Product design and marketing design share visual language but answer to different cognitive mechanisms: affordance theory and priming theory, respectively.
Sources
Norman, D. (2013). The Design of Everyday Things (revised and expanded edition). Basic Books.
Norman, D. (2008). "Signifiers, Not Affordances." Interactions, ACM, Vol. 15, Issue 6.
Gibson, J.J. (1977). "The Theory of Affordances." In R. Shaw & J. Bransford (Eds.), Perceiving, Acting, and Knowing.
Gaver, W.W. (1991). "Technology Affordances." Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (CHI '91), ACM.
Schmitt, B.H. (1994). "Contextual Priming of Visual Information in Advertisements." Psychology & Marketing, Vol. 11, Issue 1. John Wiley & Sons.
Bargh, J.A., Chen, M., & Burrows, L. (1996). "Automaticity of Social Behaviour: Direct Effects of Trait Construct and Stereotype Activation on Action." Journal of Personality and Social Psychology, Vol. 71, No. 2.
Nielsen Norman Group. "Beyond Blue Links: Making Clickable Elements Recognizable." nngroup.com.
Nielsen Norman Group. "Button States: Communicate Interaction." nngroup.com.
Nielsen Norman Group. "A Link is a Promise." nngroup.com.
Nielsen Norman Group. "Similarity Principle in Visual Design." nngroup.com.
KORTX. "7 Call to Action (CTA) Examples & How to Write Them." kortx.io.