Why this matters
A video that tests well with college-educated viewers in the top income quartile can fall flat with households earning $40,000 a year. The reverse is also true. Tone, pacing, music, on-screen talent, and messaging hierarchy all shift emotional response across income segments.
Most audience teams run surveys or focus groups, then extrapolate. Surveys tell you what people think they feel. Focus groups reward the loudest voices. Neither captures the involuntary facial micro-expressions that reveal genuine emotional reaction frame by frame.
EmotionTrac records opt-in panelists watching your video, codes their facial expressions using the Facial Action Coding System (FACS), and delivers second-by-second emotion data segmented by household income tier. You see exactly which moments trigger joy, confusion, anger, or disengagement in each group before you commit media spend.
The workflow
Step 1: Define your income tiers
Decide which household income brackets matter for your campaign. Common splits include under $35k, $35k–$75k, $75k–$150k, and above $150k. You can also test finer gradations if your product or message has a narrow target.
EmotionTrac recruits panelists who match your demographic and psychographic filters. You specify income, age, geography, education, political affiliation, purchase behavior, or any combination. The platform pulls from a national opt-in pool and screens participants to ensure clean data.
Step 2: Upload your video and set test parameters
Upload the video file (ad spot, social clip, product demo, explainer, fundraising appeal). Set your sample size per income tier. Thirty panelists per segment gives you directional data. Fifty or more per segment tightens confidence intervals.
You can test multiple versions side by side (different voiceovers, different music beds, different opening hooks) to compare performance across tiers. The platform randomizes which panelist sees which version to avoid order effects.
Step 3: Panelists watch, platform codes facial expressions
Panelists log in from home, grant webcam access, and watch your video. EmotionTrac's FACS engine analyzes their facial movements in real time: eyebrow raises, lip corners, nose wrinkles, eye squints, jaw drops. These movements map to seven core emotions (joy, surprise, sadness, anger, fear, disgust, contempt) plus neutral and confusion states.
The system timestamps every emotion shift to the exact frame. You get a continuous emotion curve for each panelist, aggregated by income tier.
Try it: Schedule an EmotionTrac demo and see second-by-second emotion tracking in action. Or visit Audience for more information.
Step 4: Review emotion curves and identify divergence points
The dashboard overlays emotion curves for each income tier on a single timeline. You see where high-income viewers spike joy while low-income viewers go neutral. You see where middle-income viewers show confusion while top earners stay engaged.
Common divergence points include pricing reveals, lifestyle imagery (luxury settings vs. everyday settings), aspiration vs. relatability messaging, and humor styles. A joke that lands with one tier can trigger contempt or confusion in another.
You also see aggregate metrics: average joy, peak negative emotion, engagement drop-off rate, and confusion spikes. These numbers tell you whether a video works broadly or only for a narrow slice.
Step 5: Edit and retest
Use the data to guide edits. If lower-income viewers disengage at the 12-second mark when you show a beach vacation scene, swap that frame for something more relatable. If higher-income viewers show contempt when you mention a discount, reframe the value proposition around quality or exclusivity.
Retest the edited version with fresh panelists from the same income tiers. Compare the new emotion curves to the original. Repeat until you hit your target response across all segments or decide to run separate creative for each tier.
What you get
You get a frame-by-frame emotion heatmap for each income tier. You know which 3-second window kills engagement for households under $50k. You know which voiceover tone resonates with six-figure earners. You know whether your call-to-action lands or confuses.
You also get comparative data across versions. If you test three different opens, you see which one generates the highest joy and lowest confusion in each tier. That clarity eliminates guesswork and internal debate.
The platform exports raw data (CSV) and summary reports (PDF) so you can share findings with creative teams, media buyers, and stakeholders. The reports include annotated video timelines with emotion overlays, making it easy to show exactly where and why a video succeeds or fails.
Common pitfalls
Testing too few panelists per tier. Twenty responses might show a trend, but you can't trust the data for high-stakes decisions. Aim for 50 minimum if budget allows.
Ignoring neutral and confusion spikes. Teams often focus on joy and anger, but sustained neutral emotion means your video isn't landing at all. Confusion spikes mean your message is unclear. Both are red flags.
Assuming one creative works for all income tiers. If you see sharp divergence in the data, running a single version wastes money. Either edit to broaden appeal or produce tier-specific creative.
Testing only finished videos. Run early-stage tests on rough cuts, storyboards, or animatics. Catching problems before final production saves time and budget.
Over-indexing on a single emotion. A video that generates high joy but also high confusion might entertain but fail to communicate. Look at the full emotional profile.