Director of Market Research Expert

Ethos; United States; Remote; USD 80,000 - 113,000; Part time

Job summary

Ethos is seeking a director of market research to help train an AI language model on professional documents, spreadsheets, and slide decks. You will design research programs, evaluate outputs, and guide go-to-market decisions with competitive landscape and pricing insights.

This is a fully remote role in the United States with flexible hours (5–20 hours per week, or more if desired). Start date ASAP; compensation is $70 per hour.

Qualifications

• 4+ years conducting market research or competitive intelligence.
• Experience building market sizing models and pricing matrices.
• Expert-level document, spreadsheet, and slide craftsmanship.

Responsibilities

• Create, evaluate, and refine AI-generated documents, spreadsheets, and slide decks.
• Develop competitive landscape analyses and market-sizing models.
• Produce clear, well-structured outputs for go-to-market and product decisions.

Skills

Market research

Competitive intelligence

Industry benchmarking

Job description

Pay

$70/hr

Compensation

$70/hour

Commitment: Flexible, 5-20 hours per week (or more if desired)

Location: Fully remote, work on your own schedule

Start date: ASAP

About this role

From the Ethos listing.
About This Opportunity
We're working with a leading foundational AI lab to find experienced directors of market research who can help train their latest language model on professional document, spreadsheet, and slide deck tasks.
We're looking for directors of market research with 4+ years conducting market research, competitive landscape analysis, or industry benchmarking to create, evaluate, and refine AI-generated documents, spreadsheets, and slide decks across core workflows: competitive landscape and positioning decks, market sizing and TAM/SAM/SOM models, win/loss and competitor battlecard analyses, industry benchmarking and trend reports, customer and stakeholder survey summaries, and pricing and feature comparison matrices.

Qualifications

• 4+ years conducting market research, competitive intelligence, or industry benchmarking analysis in an agency, consulting, or in-house strategy role
• Direct experience building market sizing models, competitor battlecards, or pricing comparison matrices used to inform go-to-market or product decisions
• Expert-level document, spreadsheet, and slide craftsmanship, with excellent written communication and attention to detail

Timeline

• 25 days of applying.
• Within 2 weeks ~30%
• Within 6 weeks ~65%
• Within 3 months ~85%

Prepare for the Ethos call

Interview Preparation

• How do you turn a vague business question into a study design with a sample, method and set of measures you can defend?
  I write down the decision the research will inform and the answer that would change it, then work backward to the smallest study that could produce that answer. That usually settles whether I need a survey, interviews, or both, and it gives me a sample size target based on the difference I need to detect rather than on a round number.
• What checks do you run on a questionnaire before it goes into the field to catch questions that will produce misleading data?
  I pilot it with a handful of respondents and ask them to think aloud, which catches double-barreled questions, leading wording, and scale labels that people read differently than I intended. I also review the order of questions, since asking about satisfaction right after a list of problems reliably pulls the scores down.
• How do you decide when a survey result is strong enough to report as a finding versus noise in the sample?
  I look at the size of the difference against the margin of error for that subgroup, and I check whether the pattern holds across related questions rather than appearing in one item alone. A difference that is significant but tiny gets reported with that caveat, and a large one in a small subgroup gets flagged as directional so nobody builds a plan on it.
• How do you analyze a set of 20 interview transcripts so the themes you report are more than the quotes you remember best?
  I code each transcript against a short framework I set before reading, then add new codes as they come up and go back to recode the earlier transcripts. Counting how many participants raised each theme, not how often it was mentioned, keeps one talkative interviewee from dominating the findings, and I pick quotes after the counts rather than before.

Common Questions

• Is Ethos regular AI training work like rating chatbot responses?
  No. Ethos is an expert