The Consumer Analyst's Staged Path From Concept Screen to Validation
On this page
A consumer analyst should stage concept research around the next funded action. Which concept deserves development, which option should remain unchanged, and which unresolved question deserves fieldwork? Subconscious can compare complete concepts through controlled experiments on simulated buyers. The useful output is a development recommendation with estimated choice effects and uncertainty, tied to actual development costs and capacity.
The Mintel comparison shows why the measured task matters. Modeled attribute-preference distributions achieved 0.79–0.88 Jensen-Shannon similarity against human research on facial moisturizer. The models used demographics alone, without preference or choice data as input. The human reference comprised two waves of 1,000 respondents each. That result supports a concrete preference-comparison use, not a percentage of correct predictions or proof that a physical product performs as described.
Fund the concept decision, not a fixed research ladder
The next research purchase should resolve an uncertainty that could change the allocation. Existing customer research may already cover the audience, complete alternatives and required outcome. Reuse sufficient evidence before commissioning another study.
Exploration can supply objections or new concepts. A controlled comparison can estimate how assigned alternatives change modeled choice. A physical-use study can establish whether a prototype meets a specified use requirement. Those are different evidence jobs. No project needs all three simply to qualify as rigorous.
For example, a verified printing fault needs a correction. A question about ink bleeding needs a test using the specified paper, ink and use conditions. Neither question is answered by a higher concept-choice estimate. Conversely, a preference decision between accurately described layouts does not automatically require a new handling study.
Compare feasible concepts before committing development capacity
Consider this explicitly illustrative planning-pad decision. The audience is adults who bought and used a paper weekly planner during the past year and still plan weekly on paper. Actual audience definitions, construction and counts must appear in the study record. The example assumes no existing study adequately compares the proposed layouts.
The current product and both new concepts are one A5 pad, 148 × 210 mm, with 52 detachable sheets of white 80 gsm paper. All have the same glued top edge and card backing. The current layout has seven equal day boxes. Concept A adds a weekly-priority area by reducing the day-box area. Concept B instead adds a weekly-spending area with the same reduction. Exact, readable artwork must show each complete layout; the concepts make no claim to improve productivity or budgeting.
Each pad costs USD12 total, including shipping and applicable tax, with delivery in five to seven business days. The hypothetical seller permits cancellation before dispatch and a full refund for an unused pad returned within 30 days of delivery, with seller-paid return postage. The seller must confirm the materials, artwork, production feasibility and all commercial terms before fielding. The outside option is to skip the pad and continue the buyer's current paper-planning practice.
Engineering and production owners supply actual artwork, setup costs, unit costs and available capacity. A concept that cannot fit the available development budget is not an actionable winner. Correct an infeasible offer before comparing preferences.
Estimate the contrast that supports the allocation
Randomly assign modeled respondent units to the current pad, Concept A or Concept B. Each arm offers its assigned pad against the identical outside option. Hold price, delivery, physical specifications and presentation conditions constant. The randomized intervention is the complete layout shown, including the space removed from the day boxes.
Estimate Concept A minus current and Concept B minus current in modeled pad-choice probability. Report actual respondent counts by arm, the estimator, and confidence intervals that account for repeated responses from the same modeled unit. Repeated answers do not create additional independent respondents. The methodology explains controlled comparisons; the study record must identify the implemented design and uncertainty calculation.
Before reading results, the product owner agrees what choice improvement would justify each concept's actual development and delivery costs. Development capacity may permit one concept, both concepts or neither. Choice probabilities do not supply sales volume, margin or a market forecast. The acceptance rule needs those commercial inputs from the responsible business owners, rather than an invented conversion from modeled choice to revenue.
Deliver a development allocation with a named accepting owner
The recommendation should identify an action and the evidence required for that action.
| Decision | Evidence needed | Accepted output |
|---|---|---|
| Develop a new layout | Its contrast and uncertainty support the prior useful-gain rule, and confirmed costs and capacity permit development | Name the layout, approved artwork, funded work and product owner |
| Retain the current pad | Adequate evidence supports no new layout meeting the agreed rule | Keep the current product and release unneeded development funds |
| Leave the allocation unresolved | Relevant uncertainty could change the decision | Name the missing comparison or input, responsible owner and cost of resolving the gap |
No observed winner is supplied in this example. If both layouts qualify but capacity permits one, compare the qualifying options under the agreed commercial rule and their direct contrast. Do not select a winner merely because one estimated gain has a smaller p-value.
Fieldwork belongs in the allocation when a required outcome remains unanswered. If the decision depends on writing comfort, the analyst should specify the physical task and acceptance criterion. If the remaining gap is a choice comparison, preserve the relevant offers and audience. Existing suitable evidence can close either gap; a new human study is not a compulsory final stage.
The final record links the evidence, exact concept versions, audience, design, estimates, uncertainty, actual costs and the accepted develop, retain or unresolved action. That record lets a product owner fund the recommendation without turning an exploratory reaction into a validated product claim.
Book a decision review. Bring the current product, feasible concepts, existing evidence, development budget and the specific outcome that could change the allocation.