Who this decision is for
- A strategy manager at an overseas company who must present evidence to leadership before Korea receives any marketing budget.
- A product manager who suspects competitors already occupy the positions the company wants and needs to see them in Korean.
- A founder without Korean-language skills who cannot read Korean search results and needs someone to interpret them accurately.
- A consultant or investor analyst who needs a marketing-side view of a company's Korea plan and wants to know what evidence is reasonable to expect.
The decision to make
The decision is what to research and how deep to go. Desk research explains the category and buyer context, search research shows the language and intent buyers use, and competitor review shows what you will be compared with. A common failure is to buy a large report that says little about how Korean buyers actually search and decide.
The trade-off is depth against cost and delay. Some uncertainty will remain however much you research, and only a live test can resolve it. The best research is targeted at the decisions ahead, such as go or no-go, positioning and channel choice, and it names what remains unknown. Use pilot testing for what desk work cannot settle.
Options compared
| Option | Best when | Watch out for |
|---|---|---|
| Desk research on category and buyers | You need context on category structure, buyer types and typical purchase paths before framing the plan. | Secondary sources may be dated or generic; check with Korean-language sources. |
| Search-behaviour research | Buyers research online and you need Korean terms, questions and result types to plan content. | Search results shift over time; treat findings as a snapshot and recheck. |
| Competitor and adjacent-brand review | You need to see how rivals present themselves, price signals, claims and content quality in Korean. | Public information is partial; do not infer private performance or sales. |
Risks to manage
- Translating rather than researching. Machine-translated English queries miss how Koreans actually phrase needs. Research in Korean with native readers.
- Over-researching. Endless study delays action. Set the questions in advance and stop when answers stop changing the decision.
- Competitor misreading. Assuming a rival's visibility equals success can mislead. Record what you observe and avoid inferring unseen numbers.
- Stale sources. Old reports may not reflect current platforms or behaviour. Check dates and confirm key points against live Korean results.
Evidence to gather first
- A list of Korean search terms and questions in your category, gathered from live results and suggestions.
- A comparison of five to ten competitors or adjacent brands, recording their Korean site, messaging, content and claims.
- Notes on which result types appear for key queries, such as blogs, shopping listings, video or local results.
- Any Korean-language reviews, discussions or community threads that reveal how buyers talk about the category.
Recommended process
- Frame the questions. SWIN and the client write the decisions the research must support, such as go or no-go, positioning and channel order.
- Desk research. We review category structure, buyer types and public context, noting dates and source quality, and mark what needs confirmation.
- Search-behaviour study. We examine Korean queries, related questions and result types across relevant search engines, and capture the language buyers actually use.
- Competitor review. We document how selected competitors present their brand, offers and content in Korean, recording observations rather than assumptions.
- Synthesis and unknowns. We deliver findings with implications for your plan and a clear list of unknowns that a pilot should test.
What SWIN can deliver
- A research questions brief linking each question to a decision.
- A Korean keyword and intent list with notes on result types.
- A competitor comparison sheet based on public Korean-language presence.
- A findings memo with implications for positioning and channel order.
- A list of open unknowns to test in a pilot.