2. Market Research

You are going to build something companies are actually struggling with — and you'll have the receipts to prove it.

Phase 2: Market Research (quick-start)

This is the fast, do-it-now version of the market research step. Follow these five steps and you will turn your target list into real intelligence and a set of problems worth building around. Do the work inside the attached research template, and use the attached doc when you want the full process behind each step.

The core idea before you start: the goal here is not a list of companies, it is a list of problems those companies actually have. Companies are just where you go to find the problems.

Step 1: Set up your tracker

Open the Phase 2 research template attached to this lesson. It has three working tabs: Company Research (one row per company), Network Tracker (people you connect with), and Signal Log (problems you spot). Drop your entire Phase 1 company list into the Company Research tab and set a Priority on each one.

Step 2: Triage first, then go deep

Do not try to deeply research a hundred companies. Two passes:

• First pass: go down the whole list, spend about 2 minutes each on the basics, and set a Priority.

• Second pass: take your top 15 to 30 and go deep, about 15 to 30 minutes each. Set Status to Done as you finish.

Depth on your best targets beats thin coverage of everything.

Step 3: Research each company, source by source

For each company on your shortlist, hit these in order and capture what you find in the sheet:

• Basics: website, LinkedIn, and Crunchbase for what they do, industry, stage/funding, size, and location.

• LinkedIn jobs: how many engineering roles are open, plus the links. Volume is a signal. Ten open roles is a very different moment than one.

• Read the job descriptions (your richest source, and the one most people skip): pull the tech stack, the team they are building, and the problems they name. Phrases like "scale our platform," "build out our AI," and "improve reliability" are direct statements of what hurts. Log those.

• News and signals: search "[company] + raises / funding / layoffs / launches / AI" and "[founder] + interview / podcast." Funding tells you they are spending and on what. Layoffs plus the stated reason tell you where they are refocusing. Launches tell you the current priority. Exec interviews often reveal the roadmap.

• Open source and engineering blog: find their GitHub org, scan active repos and open issues (a public list of their problems), and read their eng blog for stack and pain points in their own words.

• Tech stack: piece it together from job posts, the eng blog, and StackShare or BuiltWith.

Step 4: Hunt for signals as you go

This is the whole point of the phase. You are collecting problems, not filling cells. The moment you notice something that smells like a problem, log it in the Signal Log with the company, the evidence, one line on why it matters, and whether you have seen it elsewhere.

Watch for: hiring themes that repeat, pain points stated in JDs or eng blogs, open issues or gaps, topics execs keep returning to, rough edges in the product as a user, and old manual or paper-based processes no one has fixed.

Problems that show up across several of your target companies are your strongest project ideas, because solving one is relevant to all of them. And remember the logistics lesson: an obscure but real problem usually beats a flashy one, because it proves you understand the industry from the inside.

Step 5: Start your silent network

While you research, go to each target company's LinkedIn People tab and connect with the relevant engineering leaders, hiring managers, and engineers on the team. Send the connection request and nothing else. No message, no pitch. Log each person in the Network Tracker.

You are planting seeds now so that when you reach out in later phases, you are already a connection instead of a stranger. Connect now, message later, and do not spam people who are not relevant.

What you should have when you are done

• Your shortlist of 10 to 30 companies researched in depth, the rest triaged with a Priority.

• A Signal Log full of candidate problems, with the recurring ones flagged.

• A growing network logged per company.

• A clear read on which 5 to 10 companies are your strongest targets and what problems keep coming up.

That signal log and those recurring problems are exactly what Phase 3 needs. Do not move on until you can name a few real problems your target companies share.

Avoid these

• Treating this as list-building. The point is finding problems, not collecting logos.

• Going shallow on everything. Go deep on your best targets.

• Skipping the job descriptions. They are the richest signal you have.

• Not logging signals as you find them. You will forget them by Phase 3.

• Messaging your connections too early. The silent network only works if it stays silent.

• Reading only the homepage. The real signal is in the jobs, news, repos, and eng blog.

• Researching companies you do not care about. You will do it half-heartedly and miss the good stuff.

Go deeper: the full doc

This is the quick-start. The attached document walks through the exact process for reading job descriptions, gathering news, mining open source, and hunting signals, with the full breakdown behind every step. Use it whenever you want the depth.