A product completes an action, then the edit cuts to a face. What is that face actually doing? If the answer is only “showing delight,” the reaction shot becomes decorative performance: the actor is expressive, but the product is no clearer. A useful reaction shot completes a judgment for the viewer—what just happened, whom it changed, and why it matters.
The reaction is therefore not an emotion inserted after a product shot. It is the link that turns two shots into cause and effect. Where the character looks first, how long they pause, whether their body keeps moving, and when the edit returns to the product all shape what the audience believes the product has done.
A Reaction Is the Verb Between Two Shots
A 2024 PLOS ONE study from researchers at Beijing Normal University revisited the Kuleshov effect with naturalistic film material. In face-scene-face sequences, 59 participants interpreted comparable neutral faces differently depending on whether the adjacent scene was happy, fearful, or neutral. A second experiment with 31 participants reported a similar behavioral pattern. The study does not prove that a particular advertising edit will improve conversion. It does show why creators should treat a face as part of a sequence rather than an isolated image.
Imagine a desktop device arranging a pile of loose papers in an instant. Cut to a person widening their eyes and the shot tells us only that they are surprised. Show them looking at the aligned edges and then at a stopwatch, and the reaction frames speed as the unexpected result. Show them stop sorting by hand and slide the remaining papers toward the device, and the reaction becomes a decision. The emotion may be similar in all three versions, but the product meaning is not.
Decide Which Question the Character Must Answer
Before filming, temporarily remove words such as delighted, impressed, and amazed. Replace them with an observable question: Did the character notice the change? What did they understand? Did it alter what they do next?
Those questions produce three different kinds of reaction.
Attention: Show the Audience Where to Look
The character’s gaze lands on the relevant part of the product before the expression changes. This is useful when the frame contains several competing details or when the change is visually subtle.
A 2017 eye-tracking study of online banner ads found that a model’s gaze toward the product region could direct attention toward that area. The researchers also stressed the limits of the experiment, which used one product category and a specific banner format. It is not a universal rule for video advertising. Still, the study offers a practical design insight: an actor’s gaze is a compositional line.
If the actor looks into the lens before looking at the product, the viewer first receives a social address: “I am showing this to you.” If the actor keeps looking at the product, the reaction is more readily attributed to the object or event. The right choice depends on whether the scene invites the viewer into a relationship or asks them to witness a change.
Experience: Make an Invisible Sensation Legible
Scent, temperature, resistance, quietness, and comfort are difficult to explain through a product close-up alone. A change in breathing, a shoulder relaxing, a fingertip pausing, or a shift in movement speed can give the sensation a bodily basis. The objective is not a larger expression. It is a body change that unfolds on the same timeline as the product event.
When the experience touches a performance or efficacy claim, the reaction must remain a depiction of one character in one scene. It cannot replace product evidence. A relaxed character does not prove that a device is more efficient, and a comfortable expression does not establish a health benefit.
Decision: Let the Product Change the Next Action
The most commercially meaningful reaction is often a choice rather than a smile. The character sets down the old tool, gives the product a second object, passes the result to a colleague, or stops checking the outcome. The action answers why the product belongs in a life or workflow more specifically than a satisfied face can.
This also prevents the performance from stealing the scene. The face confirms; the next action delivers the meaning. For a related approach to restrained performance in absurd scenes, see Deadpan Product Ads: Directing Comedy Without Mugging.
Shot Order Determines Whether Viewers Anticipate or Confirm
Use the same footage in a different order and the reaction performs a different job.
Product action followed immediately by reaction is confirmation. The viewer receives the event first and then sees its effect on a person. It is direct and useful when a feature appears for the first time.
Reaction before the product action creates a question. The viewer wonders what the character has seen, and the next product shot supplies the answer. This can work at an opening or a second reveal, but if the reveal arrives too late, the reaction feels unmotivated.
Keeping the product action and the character in the same frame creates continuity. Gaze, gesture, and physical change no longer depend on an edit to establish their relationship. The trade-off is more demanding blocking and potentially less product detail.
A particularly useful sequence is product action, human reaction, then product proof. The final shot is not another beauty shot. It returns to the part, result, or state that changed, confirming that the character reacted to something the audience can see. This proof-back shot is valuable in short-form work because it hands the emotion back to the product.
Time the Reaction by Information, Not by Seconds
The common problem is not simply that a reaction is too long or too short. It is that the shot contains no stages. The actor jumps from neutral to a complete emotion as if they knew the result in advance, or holds for several seconds without making a new judgment.
An editable reaction usually contains at least three distinguishable moments: the gaze locks onto the change, the body receives the information, and a small decision appears. That decision may be to approach, withdraw, continue, stop, or check whether someone else noticed. The final edit may omit part of the process, but the actor should have experienced it.
For a fast rhythm, cut when the decision is just forming and let the next product shot complete it. For more aftertaste, hold briefly and allow the viewer to infer the character’s thought. Research on “cryptic reaction shots” in popular film argues that restrained, transient expressions can invite viewers to infer a character’s mental state—but only when the surrounding narrative gives them a reason to connect the shots. That research is a useful warning against treating an enigmatic face as a shortcut.
Face, Gaze, Body, and Sound Should Not Say the Same Thing
The actor opens their eyes wide, gasps, steps backward, the music rises, and someone says “That was fast.” Five signals repeat surprise. The message is not clearer; it is merely louder.
A better division of labor gives each channel a separate task. Gaze points to the cause. Breath and muscle tension establish intensity. Hands and posture reveal the next action. Sound decides when the audience notices the event. A product sound can precede the cut and make the reaction feel triggered, or arrive after a small silence and let anticipation form first.
If the product remains outside the frame, the gaze and environmental response must still indicate the same off-screen source. Otherwise the reaction becomes performance without an object. See Product Ad Framing: Creating Presence Before the Product Appears for a fuller treatment of off-screen presence.
With AIGC Characters, Build a Reaction Unit First
Reaction shots depend on fine changes in gaze, expression, and bodily continuity. When using a generated character, avoid asking one clip to perform an entire chain—notice the product, become surprised, reach toward it, and smile. Begin with a reaction unit that can be judged on its own: stabilize identity and camera position, define one eyeline, and allow only one expressive stage to change. Let editing and a separately controlled product shot establish the relationship.
This is not merely a concession to a tool. It preserves creative control. If eyeline, facial structure, or hand position drifts between adjacent shots, the viewer spends attention explaining a visual error, and the intended causal link breaks. When a real person is represented, likeness permission, permitted uses of the performance, and any required synthetic-media disclosure must be handled separately. Technical possibility is not publication permission.
Run the No-Face Test
Temporarily remove the reaction shot and watch only the product action and result. Can the viewer tell what happened? If not, the weakness may be in the product expression, not the absence of a more amazed face.
Then remove the product proof shot and watch only the character. Can the viewer explain why they reacted? If not, the performance has no reliable object.
Finally, remove the exaggerated expression while keeping gaze, pause, and the next action. Does the product meaning remain? If it does, the reaction shot has become narrative structure rather than emotional decoration.
A strong product ad does not ask an actor to testify on the product’s behalf. It lets the character locate the change, experience it, and make a choice because of it. Then the edit returns to the product and turns emotion into visible fact.
If you are planning a commercial in which product demonstration, performance, and rhythm must support the same idea, bring us the core product action, the use context, and the change the audience should understand. ONCE can design that causal chain across concept, direction, cinematography, performance, and post-production. Explore our TVC production service.
