Think Like an AI Researcher Module 5 of 6
71.4% complete
Module 05 ยท 13 min read

Commissioning AI Research from a Team

The 10 questions that separate a credible AI research brief from a vague ask. How to specify deliverables, avoid common brief failures, and evaluate the quality of what you receive.

๐Ÿ“‹ 10-question brief framework ๐Ÿ“Š Deliverable quality matrix โœ“ Research brief template

The gap between asking and receiving

Most AI research briefs fail before the research begins. A leader asks "can we benchmark these three models against our use case?" A team spends two weeks running evaluations. The results come back: a spreadsheet with accuracy numbers, no context, no recommendation, and no actionable conclusion.

This is not a team failure. It is a brief failure. The leader did not specify what "benchmark" means, what the evaluation task is, what a good result would look like, or what decision the research needs to support. The team answered the question they were asked : and the question was too vague to produce something useful.

The brief's job

A research brief does not just specify what to study. It specifies what decision the study must inform, what quality of evidence is needed to make that decision, and what the deliverable must contain for the reader to act.

The 10 questions a brief must answer

  1. What decision will this research support?

    Name the specific decision: "Proceed to pilot with Model A" or "Allocate budget for fine-tuning" or "Reject vendor proposal." Research that does not point at a decision tends to produce findings that cannot be acted on.

    Why it matters: If the team does not know what decision the research must support, they optimize for interesting findings rather than actionable ones.
  2. What is the exact task being evaluated?

    Not "how well does it do document review" but "how accurately does it identify clauses requiring legal review in commercial lease agreements, with a false negative rate below 2%." See Module 4 for the input/output specification framework.

    Why it matters: Vague task definitions produce results that cannot be compared across evaluations.
  3. What is the minimum evidence threshold to make this decision?

    Sample size, confidence level, effect size. Written before the evaluation begins : not negotiated after seeing results.

    Why it matters: Without a pre-defined threshold, the decision-maker decides whether the evidence is sufficient after seeing whether they like the answer.
  4. What are the three most likely confounders in this evaluation?

    Require the team to identify them before running the study. Prompt variation, sample selection, evaluator bias. Name each one and specify how it will be controlled.

    Why it matters: A team that has not thought about confounders before running the study will not be able to attribute results to the right cause.
  5. Who will evaluate outputs, and how?

    Automated metrics, human raters, or LLM-as-judge. If human raters: what is the rubric, what is the inter-rater reliability target (Cohen's kappa above 0.6 is a reasonable floor), and are evaluations blinded?

    Why it matters: Unblinded human evaluation with no reliability check produces numbers that cannot be trusted or replicated.
  6. Which models or approaches are being compared, and why those?

    The comparison set should be motivated by the actual decision, not by what is easy to test. If you are choosing between vendor A and vendor B, both should be evaluated under identical conditions.

    Why it matters: A comparison that excludes the relevant alternative is not useful for the decision.
  7. What data is the evaluation using, and where does it come from?

    Production data from the last 4-8 weeks is preferable to synthetic or historical data. Specify source, sampling method, and any preprocessing. Include the data preparation steps in the deliverable.

    Why it matters: Data provenance is the first thing a skeptical stakeholder will question. If the team cannot answer it, the results will not hold up in a budget meeting.
  8. What is the format of the deliverable?

    Not "a report." Specify: executive summary (one page, decision-ready), methodology section, raw results with confidence intervals, and a recommendation section that states the supported decision. See the deliverable spec below.

    Why it matters: Without format specification, the team defaults to whatever is easiest to write, which is usually a data dump rather than a conclusion.
  9. Who will review the methodology before results are collected?

    Methodology review before running the study catches design flaws that cannot be corrected after the fact. Designate a reviewer who is not on the team running the evaluation.

    Why it matters: Post-hoc methodology critiques produce conflict and delay decisions. Pre-study review eliminates the most avoidable errors.
  10. What happens if the results are inconclusive?

    Define this upfront: "If no model exceeds 80% with the required sample, we will run a second stage with a different task decomposition" or "we will delay the decision by 30 days and run the evaluation on a larger sample." Do not leave this undefined : it determines whether an inconclusive result is informative or a wasted two weeks.

    Why it matters: Inconclusive results are common. Teams without a pre-defined response tend to either declare victory (by relaxing the threshold) or declare failure (by abandoning the evaluation). Neither is the right answer.

Deliverable quality: what to require

SectionRequired contentCommon failure
Executive summary Decision supported, evidence quality, recommendation, confidence level Describes the methodology without stating a conclusion
Methodology Task specification, sample description, confounder controls, evaluation approach Lists benchmarks used without explaining why
Results Primary metric with confidence interval, secondary metrics, error analysis on failures Raw accuracy numbers with no statistical context
Limitations What the evaluation does not cover, what could change the conclusion Missing entirely : implies false certainty
Recommendation Explicit go/no-go on the decision, conditions under which the recommendation holds "Results are promising" without a clear recommendation

Red flags in what you receive

When a research deliverable lands, scan for these patterns before accepting it as input to a decision.

Red flag 1

Accuracy without confidence interval. A number like "87.3% accuracy" with no confidence interval does not tell you whether the result would hold on a different sample. Require error bars or sample size disclosure for every headline number.

Red flag 2

Recommendation missing from executive summary. A summary that describes what was done but does not state what to do next is not decision-ready. The team either does not know the answer or is hedging. Ask which it is before accepting the deliverable.

Red flag 3

Limitations section absent. An evaluation without stated limitations is overclaiming. Every real evaluation has scope boundaries. If the team did not write them, they did not think about them, which means they may have built a study that misses important failure modes.

Quality signal

A high-quality research deliverable makes you less confident in the answer, not more : because it shows you exactly where the evidence is weak and what would change the conclusion. Overconfident deliverables are a sign of incomplete analysis, not strong results.

A team delivers an AI evaluation report with 92% accuracy on your document classification task. The executive summary says "results are strong and the model appears well-suited to our use case." What is missing?
The report should include a visual chart of the accuracy trend
The executive summary should state a specific recommendation : go, no-go, or conditions : not just describe the results as "strong"
The report should compare performance across more models
The evaluation should have used a larger sample
Try this: write a one-page research brief
  1. Pick an AI evaluation your team has done recently or is planning to do.
  2. Answer the 10 questions above in writing. Each answer should be 1-3 sentences.
  3. Share the draft with the team lead who will run the evaluation. Ask: "Is anything in here ambiguous?" The questions that prompt clarification are the ones that were not answered correctly in the original brief.
  4. Review the deliverable against the five-row spec table above when it arrives.

Excited about AI, innovation, and growth?

Start a conversation
Was this module useful?
Before you go

You can now commission AI research that produces decision-ready deliverables. The final module covers communicating uncertainty: how to translate research findings into board-ready language and executive decisions.

What you learned

The 10-question brief framework, deliverable quality standards, and three red flags to scan for when a report arrives.

Module 6 preview

From Research Findings to Enterprise Decisions: how to communicate uncertainty to a board, and when evidence is strong enough to act.