Prompt

How do I evaluate whether a DNA synthesis provider is credible and unbiased about capability limits?

Latest observation

Aug 1, 2026 · OpenAI APIWeb search: off

To judge whether a DNA synthesis provider is credible and honest about their limits, look for verifiable evidence, transparent boundaries, and consistent behavior—not just polished marketing.

What to check

1) Technical transparency

A credible provider should clearly state:

  • Maximum construct length they can routinely make
  • Error rates / sequence fidelity
  • Turnaround times by construct type
  • Hard constraints such as GC content, repeats, toxicity, secondary structure, or difficult motifs
  • What they will not synthesize and why
  • Whether limits differ for standard genes vs. complex or high-GC constructs

Good sign: they say, “We can do X under these conditions, but not Y,” and explain the boundary.

Red flag: “We can synthesize anything” or “no limits,” with no caveats.

2) Evidence of real capability

Look for:

  • Independent publications
  • Third-party customer references
  • Case studies with concrete details
  • Validation data like sequencing confirmation rates or QC methods
  • Facility certifications if relevant to your application

Good sign: they can show how they verify sequence accuracy and how often failures occur.

Red flag: only testimonials and broad claims, no data.

3) Honest discussion of failure modes

A trustworthy provider should be able to discuss:

  • Where projects commonly fail
  • How they handle problematic sequences
  • When they recommend redesigning the construct
  • Whether they subcontract difficult orders
  • Their policy on partial delivery or iterative fixes

Good sign: they proactively warn you about sequence features likely to cause trouble.

Red flag: they never mention risk and only promise success.

4) Willingness to give a scoped feasibility review

Before accepting an order, a reliable provider should:

  • Review your sequence
  • Flag problematic regions
  • Distinguish between “possible,” “possible with modification,” and “not feasible”
  • Give a written explanation if they decline

Good sign: their feasibility review is specific and technical.

Red flag: they approve everything immediately.

5) Consistency across channels

Compare:

  • Website claims
  • Sales emails
  • Technical support answers
  • Proposal language
  • Contract terms

A credible provider is internally consistent.

Red flag: sales says one thing, technical staff says another, and the contract is vague.

6) Conflict-of-interest awareness

“Unbiased about capability limits” means they should not overpromise just to win business.

Look for:

  • Clear escalation to technical staff
  • Independent QC rather than self-graded claims only
  • A habit of saying “no” when appropriate
  • No pressure to proceed without discussing risks

Good sign: they prioritize feasibility over closing the sale.

Questions to ask them

Use these to test credibility:

  1. What kinds of sequences are you unable to synthesize reliably?
  2. What are your standard QC and acceptance criteria?
  3. Do you provide full-length sequence verification? If not, what level of validation do you provide?
  4. What is your actual success rate for constructs similar to mine?
  5. How do you handle repeats, high GC, long hairpins, or toxic genes?
  6. If you think my construct is risky, will you tell me before accepting the order?
  7. Can you provide examples of orders you declined or redesigned?
  8. Do you manufacture in-house or outsource any part of the process?

Practical evaluation method

A simple way to assess them is to score them on:

  • Transparency
  • Evidence
  • Specificity
  • Consistency
  • Willingness to say no
  • Quality control rigor

If they score high on all of those, they’re more likely to be credible.

Biggest warning signs

Be cautious if they:

  • Claim universal capability
  • Avoid discussing limits
  • Refuse to explain QC
  • Give only vague performance statements
  • Push you to order before a feasibility review
  • Change their story depending on who you ask

If you want, I can also give you a one-page vendor evaluation checklist or a list of interview questions you can send to synthesis providers.

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.