Consumer Preference Studies for Product Feature Prioritisation and Design
Unlock the features your customers truly value. Research Bureau designs and delivers rigorous, actionable consumer preference studies that inform product roadmaps, accelerate design decisions, and minimise costly feature missteps. Our research services translate real consumer trade-offs into ranked feature priorities, quantified value estimates, and clear design direction.
We work with product teams, UX designers, category managers, and innovation leads to convert qualitative insight and quantitative rigor into prioritized feature sets that drive adoption, satisfaction, and revenue.
Why consumer preference studies are essential for product development
Product development is full of uncertainty. Teams often face competing opinions, internal bias, and noisy feedback from early adopters. Consumer preference studies replace guesswork with statistically defensible insight.
- Reduce risk: Identify which features deliver meaningful utility and which are perceived as “nice-to-have.”
- Prioritise effectively: Allocate engineering and design resources to features with the highest predicted impact on adoption and retention.
- Design with trade-offs: Understand how consumers make trade-offs between price, features, and experience.
- Align stakeholders: Produce data-driven roadmaps that get buy-in from leadership and cross-functional teams.
Our studies create a direct line from consumer preferences to product decisions, ensuring every sprint moves the product closer to market success.
Who benefits from our service
We serve teams across sectors where feature prioritisation and design decisions matter, including:
- Consumer packaged goods (CPG) and retail product teams
- Mobile and web app product managers
- Hardware and connected devices teams
- SaaS product and UX teams
- Financial services product developers (non-medical)
- Any team launching new products, features, or packaging
If you’re deciding which features to include in a minimum viable product (MVP), which variant to launch, or how to structure tiered pricing, a well-executed preference study will save time and budget.
Our approach — rigorous, repeatable, and decision-focused
We combine industry-leading methods with structured workshops and clear deliverables. Our process is transparent and designed to produce immediate product decisions.
Step 1 — Discovery and hypothesis framing
We begin with a focused workshop to:
- Define business objectives and success metrics.
- Map assumptions and decision points that the study must answer.
- Identify target segments and competitive benchmarks.
This ensures the study directly answers product questions like “Should we prioritise offline mode over advanced search?” rather than producing generic results.
Step 2 — Method selection and study design
We select the most appropriate method(s) for the problem. Options include:
- Conjoint analysis (choice-based, full-profile, adaptive)
- MaxDiff / Best–Worst Scaling
- Discrete Choice Experiments (DCE)
- Kano analysis for delight vs baseline features
- A/B and multivariate testing for designs already in-market
- Qualitative depth interviews and usability tests to prepare stimuli
We design stimuli (feature bundles, prototypes, pricing) and construct an experimental design that maximises information while minimising respondent fatigue.
Step 3 — Sampling & fieldwork
We recruit a representative sample aligned with your target market or user personas. We handle panel sourcing, screening, quotas, and quality control.
- Online panels and screeners
- In-app intercepts for existing users
- Hybrid approaches for niche audiences
We apply care to sample size, quotas, and attention checks to ensure reliable estimates.
Step 4 — Analysis & modelling
We use advanced modelling techniques to extract actionable metrics:
- Part-worth utilities and attribute importances (Conjoint)
- Share-of-preference and market simulations
- Latent class / segmentation analysis
- Hierarchical Bayesian estimation where appropriate
- Significance testing and confidence intervals for comparisons
We translate statistical output into product implications, not just tables of numbers.
Step 5 — Actionable deliverables & workshop
We deliver a clear, prioritized set of recommendations with supporting evidence:
- Ranked feature lists with quantified impact scores
- Segmented feature preferences and personas
- Prototype recommendations and usability findings
- Market simulations for pricing and bundling
- Executive summary, full report, and raw datasets
We run a decision-focused workshop to convert insights into a feature roadmap, resource plan, and next steps.
Methods explained — how we measure preference and prioritise features
Choosing the right method depends on the objective. Below is a detailed comparison to help you understand which method is right for different product questions.
| Method | Best for | Key deliverables | Typical sample size |
|---|---|---|---|
| Choice-Based Conjoint (CBC) | Predicting market choice between multi-attribute products | Part-worth utilities, attribute importance, market share simulation | 300–1,000+ |
| Adaptive Conjoint / ACA | Efficient estimation for many attributes | Individual-level utilities, preference tracking | 200–800 |
| MaxDiff / Best-Worst Scaling | Ranking large sets of features or messages | Importance scores, clear rank-order | 200–1,000 |
| Discrete Choice Experiments (DCE) | Realistic choice scenarios with price trade-offs | Choice probabilities, willingness-to-pay | 300+ |
| Kano Analysis | Classify features as must-have, performance, or delight | Feature categories for roadmap prioritisation | 100–500 |
| A/B / Multivariate Testing | Validate live feature variants and UX changes | Conversion lift, statistical significance | Depends on traffic; power analysis required |
| Qualitative Interviews & Usability Testing | Deep exploration before quantitative testing | Design iterations, pain points, prototype feedback | 8–30 interviews |
Conjoint analysis (deep dive)
Conjoint is the gold standard for feature prioritisation when you need to estimate the relative value of multiple attributes simultaneously. It answers questions like “If we add feature X, what must we remove or price to maintain adoption?”
- We design realistic product profiles and force respondents to make trade-offs.
- Outputs include part-worth utilities for each attribute level and relative importances.
- We run market simulations to estimate share-of-preference across proposed product bundles.
- We can include price as an attribute to estimate willingness-to-pay and revenue-optimal bundles.
Example output (simplified):
- Feature A part-worth: +1.2
- Feature B part-worth: -0.3
- Relative importance: Feature A 45%, Feature B 12%
From these, we simulate that Bundle 1 (A + C) captures 38% simulated preference versus 22% for Business-as-Usual.
MaxDiff (Best–Worst)
MaxDiff provides high resolution for ranking many features. If you need a prioritized list from 20–40 candidate features, MaxDiff gives clear, actionable scores.
- Outputs a scale where higher scores = stronger preference.
- Easier for respondents than rating each feature on a Likert scale.
- Ideal when you’re choosing which features to include in an MVP.
Kano analysis
Kano helps categorise features into:
- Must-be (basic expectations)
- Performance (linearly related to satisfaction)
- Delighters (unexpected features that increase satisfaction significantly)
This helps balance investments between fixing must-haves and designing differentiators.
Statistical rigour and quality assurance
Research Bureau applies industry-standard quality controls across all studies:
- Pre-registration of analysis plan to avoid data-mining bias.
- Sample stratification and quota controls to mirror target populations.
- Attention and quality checks (speeding, straight-lining).
- Power analysis and sample-size calculation for each objective.
- Multiple imputation and robustness checks where necessary.
- Transparent reporting of margins of error, confidence intervals, and assumption sensitivity.
We provide raw data and code on request, and we’ll fully document methods so your internal analytics team can reproduce results.
Deliverables: What you’ll receive
Our deliverables are designed to be immediately actionable by product and design teams.
| Deliverable | Included |
|---|---|
| Executive summary | One-page priorities and recommended roadmap |
| Full report | Methods, results, detailed tables, and charts |
| Feature priority matrix | Ranked features with impact scores and effort axis |
| Market simulations | Share-of-preference and price sensitivity models |
| Segment profiles | Personas with feature preferences and recommended messaging |
| Raw data and codebook | CSV/JSON exports and analysis documentation |
| Interactive dashboard (optional) | Filterable insights for stakeholders |
| Workshop | 2–4 hour prioritisation workshop to convert insights into roadmap items |
All deliverables include clear “what-to-do” recommendations and design implications so engineers and PMs can act immediately.
Sample case studies (anonymised)
These case studies show how preference studies inform decisions in realistic scenarios.
Case study — Mobile app monetisation (anonymised)
- Objective: Decide which premium bundle to launch.
- Method: Choice-based conjoint with price attribute (n = 850 current users + prospects).
- Outcome: Identified a pricing tier that balanced feature uptake and ARPU, predicting a 12–18% lift in projected revenue vs. current freemium conversions. Product team adopted top-ranked bundle and phased rollout.
Case study — New hardware product feature prioritisation (anonymised)
- Objective: Prioritise features for MVP of a connected home device.
- Method: MaxDiff + Kano + prototype usability tests (n = 420).
- Outcome: Revealed three “must-have” features and two delighters that justified a small increase in BOM cost. Roadmap adjusted to include the delighters in the next release cycle, improving projected NPS.
Case study — Retail product line extension (anonymised)
- Objective: Choose packaging features for a new SKU.
- Method: Conjoint and in-store mock purchase tests (n = 1,200).
- Outcome: Conjoint predicted two packaging variants that would outperform incumbent packaging by simulated preference share. Went to pilot test, which confirmed the prediction; product team scaled the winning design.
(We can share full case documents under NDA. Contact us for an anonymised packet relevant to your sector.)
How we ensure actionable prioritisation — examples and interpretations
Turning model output into product decisions is our specialty. Below are practical examples of how results translate into prioritisation.
Example 1 — Part-worth interpretation
- If Feature X level “Yes” has part-worth +2.5 and Feature Y level “Premium” has part-worth +0.8, Feature X contributes more to choice probability.
- Decision: Prioritise engineering capacity to build Feature X before committing to Feature Y.
Example 2 — Simulated market share
- Bundle A: Features [X, Z] → 41% simulated preference
- Bundle B: Features [Y, Z] → 22% simulated preference
- Decision: Launch Bundle A as primary SKU and test Bundle B in limited channels.
Example 3 — Segment-driven roadmap
- Segment 1 (Budget Buyers): Highly price-sensitive, prefer core reliability features.
- Segment 2 (Power Users): Prefer advanced toggles and integrations; willing to pay more.
- Decision: Build core product covering Segment 1 for launch and roadmap premium features for a paid tier targeting Segment 2.
Sample-size guidance and budgets
We provide tailored pricing after scoping, but here are typical sample-size guidance and indicative timelines:
| Objective | Typical sample | Timeline (end-to-end) | Indicative cost range (ZAR) |
|---|---|---|---|
| Feature ranking (MaxDiff) | 300–600 | 3–5 weeks | 60k–150k |
| Conjoint for feature + price | 500–1,000 | 4–8 weeks | 120k–350k |
| Mixed methods (qual + quant + usability) | 200–800 + 10–30 qual | 6–10 weeks | 180k–450k |
| Rapid MVP prioritisation | 150–300 | 2–3 weeks | 45k–120k |
Notes:
- Costs vary by sample sourcing complexity, segmentation needs, and deliverable formats (dashboards and workshops add fees).
- We provide fixed-price proposals and time-and-materials options based on your needs.
- We recommend powering A/B tests with traffic-based power calculations if you plan live experiments.
Common pitfalls we avoid
We design studies to avoid frequent mistakes that lead to misleading conclusions:
- Sampling bias: We match quotas to your user base and adjust for panel effects.
- Hypothesis creep: We focus on core decisions, not exploratory fishing expeditions.
- Overloading respondents: We use efficient designs (adaptive conjoint, balanced blocks) to reduce fatigue.
- Ignoring segments: We report both aggregate and segment-level insights to avoid “one-size-fits-all” decisions.
- Misinterpreting statistical noise as signal: We apply appropriate significance testing and Bayesian shrinkage where necessary.
How to work with Research Bureau — engagement models
We offer three common engagement models, all tailored to your team’s workflow.
- Scoping and fixed-price study
- Best for well-defined objectives.
- Deliverables and timeline fixed in advance.
- Ideal for budget predictability.
- Phased program (recommended for ongoing product roadmaps)
- Phase 1: Rapid exploration and prioritisation.
- Phase 2: Quantitative validation and segmentation.
- Phase 3: Post-launch optimisation testing.
- Phased billing and iterative deliveries.
- Retainer / ongoing research partner
- Continuous insight pipeline for product teams that iterate rapidly.
- Monthly retainer covers ongoing survey waves, UX testing, and ad-hoc analyses.
We can sign NDAs and integrate with your product, analytics, and UX teams.
Example project timeline (Conjoint + Workshop)
- Week 1: Discovery workshop, objectives, and attribute selection.
- Week 2: Survey and experimental design, prototype development.
- Week 3–4: Fieldwork (data collection).
- Week 5: Analysis and simulations.
- Week 6: Deliverables, interactive dashboard, and prioritisation workshop.
Faster timelines are available for rapid MVP decisions.
Pricing transparency and ROI
We focus on measurable ROI. A small study that prevents a misguided feature launch can save multiples of its cost. We will:
- Provide a clear scope and cost estimate.
- Estimate expected precision for key metrics.
- Model potential revenue or adoption impact where applicable.
Ask us to run a quick ROI sketch for your specific product; we’ll show how study insights can pay back within your roadmap planning cycle.
FAQs — quick answers to common questions
Q: How do you ensure results reflect real-world behaviour?
A: We use choice-based designs and include price trade-offs and realistic scenarios. Where possible, we validate with in-market A/B tests.
Q: Can you segment results by demographic and behavioural groups?
A: Yes. We routinely run latent class and cluster analyses to produce segment-specific recommendations.
Q: How is privacy handled?
A: We comply with local data protection regulations. Respondent data is anonymised for reporting unless explicit consent is given.
Q: Do you provide raw data?
A: Yes. Full datasets and codebooks are available on request.
Q: Can you work with internal analytics teams?
A: Absolutely. We collaborate with in-house teams and can hand over models, code, and workshops to embed learnings.
Example outputs (visual and numeric) — excerpts
Below are examples of the types of outputs you can expect (presented here in simplified form).
Part-worth utility excerpt (Conjoint)
| Attribute | Level | Utility |
|---|---|---|
| Battery life | 24 hours | +1.8 |
| Battery life | 48 hours | +3.4 |
| Offline mode | No | -1.2 |
| Offline mode | Yes | +1.2 |
Feature priority (MaxDiff excerpt)
| Feature | Score |
|---|---|
| Sync across devices | 82 |
| Offline mode | 76 |
| Advanced search | 55 |
| Customisable themes | 22 |
Market simulation (simplified)
| Bundle | Simulated preference |
|---|---|
| Core + Sync | 44% |
| Core + Offline | 28% |
| Core only | 18% |
| Premium bundle | 10% |
These outputs are followed by explicit design and roadmap recommendations in our final report.
Ready to prioritise features with confidence?
Share a brief with us and we’ll provide a tailored proposal and quote. To get started:
- Fill out the contact form on this page and attach any product briefs or feature lists.
- Click the WhatsApp icon to message us directly for a quick chat.
- Email us at [email protected] with a summary and any timelines or budgets.
We usually respond within one business day and can schedule an initial scoping call to define objectives and recommended methods.
Why Research Bureau?
- Deep product research expertise across digital, hardware, and consumer goods.
- Methodological rigour combined with practical product thinking.
- Clear, decision-focused deliverables that drive product roadmaps.
- Flexible engagement models for one-off studies or ongoing partnerships.
We partner closely with product teams to ensure insights are acted upon and measured in subsequent releases.
Final notes — what to prepare before contacting us
To speed scoping and quoting, please provide:
- A brief description of the product or feature set.
- Your target market and any known user segments.
- Key decisions you need the study to inform.
- Preferred timeline and budget (if known).
- Any existing research, prototypes, or competitive benchmarks.
Send this information via the contact form, WhatsApp, or email [email protected] and we’ll return a scope and costing within 48 hours.
Research Bureau helps product teams make fewer bad bets and build features customers actually want. Share your product brief today and let’s prioritise with confidence.