Keeping a Sponsor's Terminology Consistent Across a Filmed Course
Classify each term as kept-in-English or rendered-in-Korean, record it once, and check narration, captions, lab code, and the listing against that single glossary.
The buyer question
How does Hong keep a sponsor's product terminology consistent across a filmed course's narration, captions, lab, and listing?
The output here is a terminology consistency map built for one sponsored course, tying every product term to a single kept-English or rendered-Korean decision. Assign the decision to the concept, not to a single sentence, record the approved rendering and any retained English form once, and then check the narration script, on-screen captions, the lab's code comments, and the Inflearn listing copy against that same entry. Hong recommends treating the glossary as a governed production artifact for the course, reviewed with the sponsor's technical reviewer, rather than a style choice left to whoever is writing a given scene.
Reading the decision in context
What this decision actually asks of the team.
One glossary, built for one course
Each sponsored course gets its own terminology consistency map, tied to that product's specific terms rather than a generic house style. This keeps the glossary accurate to the actual product rather than reused loosely from a previous, unrelated engagement.
The map assigns a decision to each concept once, with the approved rendering, the retained English form, and disallowed variants recorded together, so a term is not reopened and reinterpreted every time a different person on the production team writes a new scene.
Four surfaces, one decision per term
A viewer encounters a course's terminology in four places: the spoken narration, the on-screen captions or slides, the lab's own code comments, and the Inflearn listing copy they read before enrolling. Each of these is checked against the same glossary entry rather than translated independently.
Executable identifiers inside the lab's code, field names, commands, and parameters, are the one category kept unchanged regardless of the glossary's Korean-rendering decisions elsewhere, so a student can still copy and run the sample exactly as shown.
Sponsor review keeps naming decisions accountable
Product-specific terms go through the sponsor's named technical reviewer before filming locks the script in. This matters because a plausible-sounding Korean rendering can quietly denote something different from what the sponsor actually means by a given term.
That review is scoped to the terms themselves, not to editorial phrasing choices elsewhere in the narration, which stay with Hong's production team as the people responsible for how the course actually teaches the material.
The glossary is a production discipline, not an outside authority
The linked GraphQL and distributed-tracing pages are technical artifact examples from Hong's own unsponsored catalog, showing how he keeps terminology connected across an explanation. They are not evidence of any sponsored course's terminology work, since none has published yet.
A consistency map earns its value by keeping one course's language coherent for its own narration, captions, lab, and listing. The sponsor's technical reviewer remains the authority on what a product's own terms actually mean.
The framework
Terminology Consistency Map
Hong recommends building this map before scripting starts, so the same product concept does not get three different names by the time narration, captions, lab comments, and listing copy are all written by different people on the production team.
- The product's own terminology as used in its documentation and UI
- A draft list of terms likely to appear in narration, captions, and the lab
- The lab's code samples and their existing comments
- Draft Inflearn listing copy
- A named sponsor-side reviewer who can approve or correct a term
Classify each term as kept or rendered
Sort each product term into keep-in-English, render-in-Korean, or render-with-English-in-parentheses. API identifiers, product names, and established technical tokens usually stay in English; conceptual and task language often renders more clearly in Korean.
Define one entry per concept
For each term, record the source word, the approved Korean rendering when used, the retained English form, a short usage note, and disallowed variants. The entry belongs to the product concept, not to a single sentence in the script.
Align the narration script with on-screen captions
Confirm the word spoken in the Korean narration matches the word shown in any on-screen caption or slide for the same concept. A viewer hearing one term while reading another loses confidence in both.
Check the lab's code comments and the Inflearn listing
The lab's code comments and the Inflearn listing copy are terminology surfaces too. Confirm both use the same approved term as the narration, and keep executable identifiers in the lab's code unchanged regardless of any Korean rendering used around them.
Review the glossary with the sponsor's technical reviewer
Have the sponsor's named technical reviewer confirm product-specific terms before filming locks in. This keeps naming decisions accountable to the sponsor rather than to the production team's best guess.
- Whether the underlying product is a fit for the Korean VOD course format at all belongs in the fit-check guide.
- How the glossary fits into the overall production pipeline belongs in the production guide.
- The lab's own design, separate from its terminology, belongs in the lab design guide.
- Grading Hong's production capability from the unsponsored catalog belongs in the evidence-read guide.
Failure modes
Where this approach should stop or narrow the work.
An established English term gets translated anyway
A well-known API identifier or product name is rendered into Korean or transliterated, and a viewer can no longer connect the narration to the product's actual interface. Mark these terms as retained English and treat any Korean rendering as an error.
The narration and the on-screen caption disagree
The spoken word and the caption on screen use different terms for the same concept, which reads as a production mistake rather than a deliberate stylistic choice. Check both against the same glossary entry before filming.
Executable identifiers in the lab get localized
A field name or command inside the lab's sample code is rewritten to look more familiar, which breaks copy-paste for a student following along. Keep executable identifiers unchanged and localize only the surrounding comments.
The Inflearn listing uses different terms than the video
A viewer who reads the listing before enrolling and then watches the video encounters two different vocabularies for the same product. Check the listing copy against the same glossary used for narration and captions.
Questions on this guide
Frequently asked about this decision.
Which terms stay in English in a Korean-language course?
API identifiers, product names, and established technical tokens a viewer would search or type usually stay in English. Conceptual and task-describing language is often rendered into Korean when a natural equivalent reads more clearly, and the category rule is written down rather than decided sentence by sentence.
Does the sample code in the lab get translated into Korean?
No. Executable identifiers, field names, commands, and parameters stay exactly as the product defines them so a student can copy and run the sample without it breaking. Only the surrounding comments and narration are candidates for a Korean rendering.
Who has final say on how a specific product term gets rendered?
The sponsor's named technical reviewer confirms product-specific terms before filming locks the script in. Hong's production team owns the general terminology discipline and how it applies across narration, captions, and the listing.
Does every sponsored course get its own terminology map, or is there one shared glossary?
Each sponsored course gets its own map built around that specific product's terms. A shared house style would risk mismatching a sponsor's actual product vocabulary, so the glossary is scoped to the engagement it serves.
Apply this recommendation
Share your product URL for a bounded Korea-facing next step.
Hong can use the product surface, current documentation, target evaluator, and Korea goal to recommend a practical first asset without implying official distribution or guaranteed adoption.
No Gmail? Open in your mail app or write to hong@koreafirstrun.dev.