Prompt

How do I evaluate whether a design research consultancy is credible and unbiased for industry-specific research?

Latest observation

Jul 27, 2026 · OpenAI APIWeb search: off

To evaluate whether a design research consultancy is credible and unbiased for industry-specific research, I’d look at it in three layers: expertise, methodology, and incentives. Here’s a practical checklist.

1) Check industry credibility

Ask whether they truly understand your sector, not just “design research” in general.

Look for:

  • Relevant past projects in your industry or adjacent ones
  • Case studies with clear problem statements, methods, and outcomes
  • Familiarity with your market’s regulations, users, workflows, and jargon
  • Evidence they’ve worked with similar stakeholder types: operators, consumers, clinicians, engineers, etc.

Good sign: They can talk concretely about industry constraints and tradeoffs.
Red flag: They rely on broad design language but stay vague on domain specifics.

2) Examine their research methodology

A credible consultancy should be able to explain exactly how they reach conclusions.

Ask:

  • What methods do you use and why?
  • How do you recruit participants?
  • How do you avoid sampling bias?
  • How do you triangulate findings across methods?
  • How do you separate observation from interpretation?

Good sign: They can justify method choice based on your research question.
Red flag: They use the same “proprietary process” for everything without explaining tradeoffs.

3) Evaluate bias and independence

You want to know whether their conclusions are shaped by client preferences or hidden commercial incentives.

Ask:

  • Do you have any vendor, tool, or implementation relationships that could influence findings?
  • How do you handle situations where findings contradict a client’s assumptions?
  • Do you present alternative hypotheses or just one preferred narrative?
  • Can we see raw themes or evidence behind the recommendations?

Good sign: They distinguish clearly between evidence, inference, and recommendation.
Red flag: They seem eager to confirm what you already believe.

4) Review their deliverables

Strong consultancies make their evidence traceable.

Look for reports that include:

  • Sample size and participant profile
  • Recruitment criteria
  • Research limitations
  • Direct quotes, observations, or artifacts supporting claims
  • Confidence levels or strength of evidence
  • Clear separation between findings and design recommendations

Good sign: The report makes it easy to audit how they got to each conclusion.
Red flag: Polished insights with little supporting data.

5) Assess the team, not just the firm

Sometimes the company looks strong, but the actual team assigned may not be.

Ask:

  • Who will actually do the work?
  • What is each person’s industry background?
  • Who on the team has methodological expertise?
  • Will senior staff be involved in analysis and synthesis?
  • Can we speak with the actual project lead?

Good sign: The named team has both domain and research depth.
Red flag: Sales team is impressive, but execution team is opaque.

6) Look for transparency about limitations

Credible researchers don’t overstate certainty.

Ask whether they:

  • Acknowledge sample limitations
  • Distinguish between exploratory and confirmatory work
  • Note when findings may not generalize
  • Explain when evidence is weak or mixed

Good sign: They are comfortable saying “we don’t know yet.”
Red flag: They present every insight as definitive.

7) Check reputation, but verify it

References matter, but use them carefully.

Do:

  • Speak to former clients about process, rigor, and neutrality
  • Ask whether the consultancy challenged assumptions
  • Ask whether recommendations were evidence-based and actionable
  • Check whether results held up after implementation

Be careful: Testimonials can be curated. Try to speak with at least one reference who had a difficult project or contested findings.

8) Watch for commercial alignment issues

Bias can come from business model, not just the researchers.

Examples:

  • They sell design services and may steer research toward solutions they can implement
  • They partner with vendors whose products align with their recommendations
  • They reuse templates or prior frameworks in ways that fit their own offerings

Ask: “What would you recommend if the answer is not a design intervention?”
A credible consultancy should be willing to recommend non-design actions if the evidence points there.

9) Run a small paid pilot

If the project is important, start with a smaller phase.

For example:

  • A discovery sprint
  • A recruitment and discussion guide review
  • A pilot interview round
  • A short methods audit

This lets you evaluate:

  • Clarity of thinking
  • Participant handling
  • Analytical rigor
  • Responsiveness to critique
  • Whether they remain objective when findings are uncomfortable

10) Use a simple scorecard

You can score each consultancy from 1–5 on:

  • Industry expertise
  • Methodological rigor
  • Transparency
  • Independence / conflict management
  • Quality of evidence in deliverables
  • Team fit
  • Responsiveness to critique
  • Reference quality

If one firm scores high on polish but low on transparency or independence, be cautious.


Short version: what credible and unbiased looks like

A strong consultancy:

  • Understands your industry deeply
  • Explains its methods clearly
  • Shows evidence for conclusions
  • Separates facts from interpretation
  • Discloses conflicts or incentives
  • Admits uncertainty
  • Can be challenged without becoming defensive

Biggest red flags

  • Vague about methods
  • Overconfident conclusions
  • No raw evidence or participant detail
  • No discussion of limitations
  • Hidden commercial relationships
  • “Research” that seems designed to sell a predetermined solution

If you want, I can turn this into a vendor evaluation checklist, a RFP question set, or a scorecard template you can use to compare consultancies.

Observation history

Unlock observation history

Create a free Obsurfable account to unlock historical responses, model comparisons, and deeper brand intelligence.

How did Obsurfable measure this prompt?

Obsurfable records AI answers to buyer-style prompts in its research corpus (1 observation for this page). Metrics are distributions over observations, not a single static ranking.

Which AI systems does Obsurfable collect answers from?

OpenAI, ChatGPT, Google, Gemini, Google AI Mode, Anthropic, Claude, Perplexity, Grok, DeepSeek, Mistral, Copilot, and Meta AI.