A 60-minute deep research workflow that produces a full competitive landscape, customer insights, and a 10-page brief — used by founders, consultants, and PMs in 2026.
A market research project that used to take a junior consultant a full week now takes a focused person about an hour. Not because AI is magic, because the bottleneck for research has always been searching, reading, and synthesizing. AI handles all three at speed and quality that would have been an entire team in 2022.
This is the exact 60-minute workflow we use to brief on a new market: B2B SaaS verticals, consumer categories, geographic markets, regulatory landscapes, anything where you need to walk into a room knowing more than the people who hired you.
A 10-page brief covering:
Optional but powerful: a $20 month of access to a transcript site (Tegus, AlphaSense) for primary source interviews if your market has them. Worth it for B2B research where customer language matters.
This is the step that makes or breaks the next 50 minutes. Vague questions get vague briefs. Sharp questions get sharp briefs.
Bad: "Research the meal kit market."
Good: "Brief me on the US meal kit market for working parents of young kids. I want to know who the top 8 players are, what the average pricing per meal looks like, what reviews say about portion sizes and convenience, and where this market is heading as more grocery delivery services add prepared meals."
Write your question that precisely. Five sentences max, but every sentence should narrow the scope.
In Perplexity, switch to Pro Search and run these eight queries (replace bracketed terms with your market):
For each query, Perplexity returns a sourced summary with citations. Copy the response and the citation list into a Google Doc, one query per section. You now have ~15-20 pages of sourced raw research.
Pro tip: when a citation looks high-quality (industry report, primary source, well-known publication), open it and skim. Don't trust Perplexity's summary blindly for the most important claims, verify the top 5-10 numbers in the original sources.
Open NotebookLM, create a new notebook, and add as sources:
Now ask NotebookLM the synthesis questions:
NotebookLM answers with explicit source citations, so you can immediately verify claims. This is the unfair advantage of NotebookLM over a regular LLM, it tells you exactly which source said what.
Save its key answers into your Google Doc.
In NotebookLM, click Generate Audio Overview. It produces a 12-15 minute conversational podcast where two AI hosts walk through your sources. Let it run in the background.
Why this matters: listening to a synthesized version of your own research catches things your eyes miss. Sometimes the best insight in your brief is a connection between two sources that you only noticed when you heard the AI hosts say it back.
You'll listen to this on your walk after the hour is done. It often reshapes the recommendations section.
Open Claude. Paste in:
Prompt:
Draft a 10-page market research brief based on the research below. Structure: Executive Summary (1 page), Market Overview with size and growth (1 page), Competitive Landscape with positioning matrix (2 pages), Customer Insights from reviews (2 pages), Trends and Where the Market is Going (1 page), Strategic Recommendations (2 pages), Sources Cited (1 page). Cite sources inline using [1], [2] format. Use direct, executive-friendly language. Avoid filler.
Claude produces the draft. Read it once for structure. Then go back and:
Add 30 minutes if your market has primary-source interviews you want to include from a transcript service.
Once you have a brief, the work shifts:
The ability to do a credible market brief in an hour fundamentally changes how often you can do them. Most teams research a market once at founding and never again. With this workflow, you can re-brief quarterly. The competitor you missed in Q1 won't surprise you in Q3.