The research gap

The "interesting but wouldn't buy" problem is structural, not methodological.

The scale of the problem: Approximately 80–90% of new consumer goods launches fail within 12 months of launch (Nielsen, Cambridge Innovation Advisors). The majority are preceded by research that predicted success. The failure does not happen at market — it happens in the research room, where the methodology cannot detect the difference between genuine desire and socially acceptable approval.

Every FMCG research team has experienced the same failure mode: a concept tests brilliantly. Focus group participants call it interesting, novel, distinctive. Quantitative scores are strong. The brand team greenlights. The product launches. Sales disappoint.

The failure does not happen at market — it happens in the focus group room, the moderated discussion, the post-session recall survey. When participants say a concept is "interesting," they are producing the safest social response available. Interesting is positive, inoffensive, and non-committal. It is also not a purchase signal.

The genuine emotional response — the one that actually predicts whether someone reaches for a product on a shelf — is involuntary, pre-verbal, and occurs within 200 milliseconds of first exposure. No focus group moderator, survey question, or post-session recall exercise can capture it. FACS analysis can.

EchoDepth measures VAD (Valence, Arousal, Dominance) continuously from the moment stimulus exposure begins. The result is a emotional response profile for every concept, every participant, every second of exposure — before any social moderation has occurred.

Traditional research — what you get

Social desirability bias — consumers describe concepts as interesting because it is the safest group response

The articulation gap — first-exposure emotional reactions cannot be accurately described in words

Concept contamination — carryover between concepts in sequential tests distorts individual responses

Group dynamics — dominant voices shift group opinion; minority responses are suppressed

End-of-session rating — you get a number at the end, not the signal throughout

EchoDepth — what you get instead

Involuntary signal — facial Action Units cannot be consciously controlled; they reveal genuine response

First-exposure capture — emotional data recorded at the moment of exposure, not recalled afterwards

Individual, not group — every participant analysed independently; no peer contamination

Desire vs curiosity distinguished — VAD scoring separates genuine purchase signal from polite interest

Moment-to-moment timeline — see exactly when and where emotional response peaks or drops

Applications

Every stage of the FMCG research cycle. Every type of stimulus.

Packaging

Packaging concept and design testing — packaging research software

Test pack concepts against each other and against existing range at first-exposure. EchoDepth identifies which design drives genuine positive affect, which drives neutral curiosity, and which generates suppressed aversion — broken down by target demographic, shelf position, and colour/texture variant. Carryover analysis detects concept contamination in sequential tests.

Brand

Brand identity and repositioning evaluation

Measure the involuntary emotional response to brand expressions — logo evolution, new design language, updated pack architecture — before committing to a repositioning. EchoDepth identifies which brand elements carry positive emotional equity and which generate confusion or indifference that stated preference questions systematically miss.

NPD

New product introduction research

At first exposure to a new product — name, visual, proposition — EchoDepth captures the emotional response before any rationalisation occurs. The result is the only reliable signal of genuine concept resonance: the pre-verbal feeling that determines whether someone reaches for a product or walks past it. Target audience segmentation shows which demographic the concept genuinely lands with.

Advertising

Advertising and creative pre-testing

Second-by-second emotional curves across video creative — showing exactly which moment lands, which triggers concern, and where attention drops — before the campaign goes live. Post-campaign emotional shift measurement shows whether the campaign has genuinely moved brand emotional response or merely increased recognition. Audio, voiceover, and music are analysed independently to identify the driver.

Retail

Shelf presence and retail environment

Simulated shelf configurations delivered remotely. EchoDepth measures attention direction, emotional response to individual SKUs or brand bays, and the emotional state at the moment of simulated decision. Captures the genuine in-store physiological response that determines purchase behaviour, without facility cost or geographic constraint.

Price

Price point emotional response

Price sensitivity research using direct questioning produces systematically biased results — consumers always report more sensitivity than they express behaviourally. EchoDepth captures the involuntary valence and dominance response at the moment of price exposure, distinguishing genuine reluctance from automatic social negotiation. The result is a realistic price elasticity signal, not a stated preference distortion.

In practice

What EchoDepth reveals that traditional FMCG research cannot.

Scenario: three-pack concept test

Three packaging concepts tested sequentially with 40 participants. Stated preference at end of session: Concept C wins (voted "most interesting" by 52%). Self-report scores are similar across all three.

EchoDepth VAD output at first exposure:

Concept A
V 0.78 ↑
Concept B
V 0.34 ↓
Concept C
V 0.65 →

Insight: Concept A generates the highest genuine positive valence at first exposure. Concept C was voted "most interesting" — but "interesting" and "desire" are different emotional states. Concept A launches. Concept C becomes a future variant.

Scenario: advertising pre-test

A 30-second TV ad tested with 60 target consumers before broadcast. End-of-session recall scores are positive. Brand tracking shows 71% recall intent. Standard pre-test: pass.

Seconds 0–8 (brand opening): V 0.71, A 0.62 — engaged, positive

Seconds 9–16 (price reveal): V drops 0.71 → 0.38. AU4 brow tension. Arousal spike.

Seconds 17–24 (product hero): V recovers to 0.55. Dominance low (0.31).

Seconds 25–30 (brand close): V 0.61. Dominance still below baseline.

Insight: Price reveal triggers suppressed negative response that self-report masked. The ad recovers emotionally but dominance (sense of control and confidence) does not fully return. Recommendation: reorder to reveal price after the product hero. Recut tested — dominance holds through to brand close.

Scenario: segment divergence in brand repositioning

A heritage FMCG brand tests a modernised visual identity with two segments: loyal existing buyers (35–54) and target new buyers (22–34). Group sessions report that both groups like the new design.

Existing buyers (35–54): V 0.42, A 0.71 — high arousal, moderate valence. Nostalgia disruption signal.

Target new buyers (22–34): V 0.74, A 0.58, D 0.69 — strong positive, confident. Genuine desire signal.

Insight: The new design wins the target segment decisively. But existing buyers — who said they liked it — show a disruption response that predicts brand switching risk. Recommendation: phased rollout with transitional pack architecture retaining two heritage cues. Loyalty protected without sacrificing new buyer appeal.

Scenario: NPD concept elimination

Five NPD concepts tested. All five score between 6.2 and 7.1 on standard appeal questions. The research team cannot differentiate on stated preference alone. The brand team needs to take two concepts to the next stage.

EchoDepth first-exposure emotional ranking (genuine desire signal):

1st — Concept D: V 0.81, D 0.74

2nd — Concept A: V 0.71, D 0.68

3rd — Concept C: V 0.63, D 0.51

4th — Concept E: V 0.58, D 0.44

5th — Concept B: V 0.41, D 0.28 (concealed indifference)

Insight: Stated preference gives a 0.9-point range across five concepts. EchoDepth gives a 0.4-point valence range and a 0.46-point dominance range. Concepts D and A go to the next stage. Concept B — which scored 6.5 on stated appeal — is eliminated on the basis of a concealed indifference signal that no standard metric would have detected.

Output

From emotional capture to marketing recommendation.

Every EchoDepth FMCG engagement delivers a structured report designed for brand teams, innovation leads, and agency partners — not a data file that requires interpretation.

EchoDepth consumer research workflow: Interviews (emotional coding, audience emotion map), A/B Testing (live variant vs emotional score), Product Testing (first-use friction, device UX signal), Competitor Analysis (rival campaign emotion, positioning opportunity) — all feeding into continuously revised Marketing Copy.

EchoDepth consumer research workflow — from emotional capture through to marketing recommendation.

Concept ranking

Genuine emotional response ranking for each concept tested, with first-exposure VAD profiles and segment breakdowns. The insight no stated preference question can provide.

Emotion timeline

Moment-to-moment VAD curve for each stimulus — showing exactly when the emotional response peaks, drops, or shifts during exposure. For advertising: second-by-second.

Segment comparison

How different target audiences — by age, loyalty, category usage — respond to the same stimulus. Where they converge and where they diverge, with statistical confidence.

Recommendation

A specific recommendation for each research question: which concept to advance, which creative to cut, which pack element to revise, and why — anchored in the emotional data.

FMCG research with EchoDepth Insight

Frequently asked questions.

Why does traditional FMCG consumer research produce distorted data?
Because the research context itself is the distorting force. In focus groups and moderated interview settings, participants produce the safest social response available — "interesting" or "I like it" — rather than expressing genuine purchase intent or aversion. Social desirability bias is amplified in group settings by peer dynamics. EchoDepth captures the involuntary physiological response before any social moderation has occurred.
How does EchoDepth distinguish genuine desire from polite interest?
Through the combination of valence and dominance in the VAD model. Genuine desire produces high valence (positive affect) paired with high dominance (sense of confidence and agency) — the consumer feels good about the concept and feels in control of the decision. Polite curiosity produces moderate valence paired with moderate to low dominance — the consumer is interested but not yet convinced. This distinction is invisible to self-report but measurable in the facial Action Unit combination.
Can EchoDepth test concepts in sequential order without carryover?
Yes, and EchoDepth specifically analyses carryover effects in sequential concept tests. The first-exposure emotional response for each concept is captured independently. The platform identifies when a participant's emotional state from concept 2 is influencing their response to concept 3, and flags this in the analysis. This is one of the most commercially valuable outputs — identifying which concepts genuinely stand alone and which are dependent on sequence position for their apparent appeal.
How many participants do we need for statistically reliable FMCG research?
For concept ranking studies comparing 3–5 concepts with a single target segment, EchoDepth produces reliable directional findings with 20–30 participants per segment — comparable to qualitative research sample sizes, but with quantified emotional output. For studies requiring statistical significance across multiple segments or longitudinal tracking, larger samples are agreed at the research design stage. EchoDepth remote delivery makes recruiting larger, more diverse samples significantly more achievable than facility-based research.
What is the turnaround time for an FMCG research report?
For a concept testing study of 20–40 participants, a full EchoDepth report is typically delivered within five to seven working days of session completion. For advertising pre-testing studies, delivery is within three to five days. Session setup and participant briefing are completed within five working days of engagement agreement.

Client perspective

"Using EchoDepth from Cavefish allows us to validate ideas quickly, minimising the risk of launching a product or idea."
GT

Gethin Thomas

CEO, Iterate

Related reading

Methodology

Why focus groups fail to predict market success

The three structural problems that make focus groups unreliable — and what to use instead.

Read insight →

NPD & concept testing

How to validate a product concept before launch

Capturing genuine purchase signal rather than approval — in 10–15 working days.

Read insight →

Get started

Your next FMCG concept test. With emotional truth included.

Book a 30-minute discovery call. Bring your current research challenge — packaging test, brand repositioning, advertising pre-test, or NPD — and we will show you exactly what EchoDepth adds to what you already run.