Key takeaways
- Start from a decision you need to make, such as entering a new industry or pricing a new model, and research only what that decision needs.
- Free public sources in Korea cover a lot: national statistics, trade data, public procurement and company disclosures.
- Ten customer interviews usually teach more than a survey nobody answers. Your own quote and inquiry records are the most underused data you have.
- AI assistants are now part of the market. Check which suppliers they name for your category and which sources they cite.
Why do it yourself
Small manufacturers rarely have the budget for a full research project, and a generic industry report often answers questions nobody in the company was asking. Research done in-house is narrower, but it is aimed at your product, your customers and the decision on the table.
It also builds knowledge that stays in the company. The salesperson who interviews ten customers keeps what they learned and uses it in the next quote, which a report on a shelf cannot do.
Step 1: start with a decision, not a survey
Write down the decision the research must support and the two or three questions that would change it. For example: should we offer our packaging machine to cosmetics plants as well as food plants? The questions might be how many such plants exist, what they use today and what they would need to switch.
- Market size: how many companies could buy, and roughly how much.
- Customer need: what problem they have and how they solve it today.
- Competition: who already serves them and how they are described.
- Access: which channels, distributors and events reach them.
If a question would not change the decision, drop it. This keeps the project to weeks rather than months.
Step 2: use free public data
Korea has a good range of free sources for industrial markets. None of them gives a finished answer, but together they let you estimate the size of a segment and spot who is active in it.
| Source | What you can learn | Good for |
|---|---|---|
| KOSIS national statistics portal | Number of companies and employees by industry and region | Sizing a target segment |
| Korea Customs Service trade statistics | Import and export volumes by product code and country | Spotting import competition and export demand |
| KONEPS public procurement (Nara Marketplace) | Tenders and awarded contracts for your product type | Finding public buyers and price ranges |
| DART electronic disclosure | Revenue, business description and risks of listed and audited firms | Profiling customers and competitors |
| KOTRA overseas market news | Country and industry briefs written by trade offices abroad | Early checks on export markets |
| Industry associations | Member lists, event calendars, sometimes industry statistics | Building a prospect list |
Write down the source and date next to every number you use. When someone asks where a figure came from six months later, you will be glad you did.
Step 3: interview ten customers
Surveys sent to busy plant managers rarely come back. A twenty-minute call or a visit usually works better, and ten of them are often enough to see the pattern. Include a few customers you lost and a few companies that never bought from you.
- What happened that made you start looking for a solution?
- How did you solve it before, and what was wrong with that?
- Where did you look for suppliers, and did you ask an AI assistant?
- Who else in the company was involved in the decision?
- What almost made you choose someone else?
- If you could change one thing about products like ours, what would it be?
Ask about past events, not opinions about the future. "What did you do last time" gives you facts. "Would you buy this" mostly gives you politeness.
Step 4: mine your own sales records
If your records live in several places, such as a CRM, email folders and a salesperson's notebook, do not wait for a perfect system. Export the last two years into one sheet with a few columns: date, company, industry, channel, product, result and reason. An afternoon of cleanup is usually enough to start.
Most manufacturers sit on years of useful data they never analyse: inquiry logs, quotes, won and lost deals, sample requests and after-sales calls. Even a simple spreadsheet export can answer questions that no outside source can.
- Which industries send the most inquiries, and which of them actually order?
- At which step do quotes usually stall, and what reason is given?
- Which channel brought the customers who reorder: trade shows, distributors or web search?
- Which questions come up again and again in calls and emails?
The last list, the repeated questions, is also raw material for your FAQ and product pages, and for the questions buyers are likely to ask AI.
Step 5: use trade shows as a research trip
A trade show is the cheapest place to see a whole market in one day. Before you go, prepare a short list of what to observe: which competitors exhibit, what they put on the front of the booth, which applications they demonstrate and what visitors ask them.
Collect brochures and price lists where they are offered, take notes the same evening and share them with the team within a week. The value fades quickly if the notes stay in one person's bag.
Step 6: research how AI describes your market
AI assistants are useful research helpers: they can summarise a regulation, list application areas or suggest interview questions. Treat their output as a starting point and verify every fact against a primary source before you use it.
They are also part of the market you are researching. Ask ChatGPT, Gemini and Naver AI Briefing the questions your target buyers would ask, and record which suppliers are named, how they are described and which websites are cited. That shows you who buyers meet first, before they ever reach a supplier's site.
Doing this by hand works for a handful of questions. Prompt Demand shows which questions buyers in your category ask AI, and GEO Insight tracks which companies and sources the answers mention over time.
A three-week research plan
| Week | Work | Output |
|---|---|---|
| Week 1 | Write the decision and questions, collect public data, analyse your own records | Estimated segment size and inquiry patterns |
| Week 2 | Run ten interviews, check AI answers for the buyer questions | Interview notes and a list of suppliers named by AI |
| Week 3 | Summarise findings, discuss with sales and management, decide | A two-page summary and the decision |
Keep the final summary to two pages: what we wanted to decide, what we found, what we recommend and what we still do not know. Honest gaps are more useful than confident guesses.