Last week, my partner had to find a good translation into Spanish for "wireless hot spot". A first search in the
Microsoft language portal suggested "zona interactiva inalámbrica" (2 hits), while a search in the KudoZ glossaries gave "zona local de cobertura" (1 hit only), although, of course, we found plenty of results for "hot spot" and for "wireless" not combined in the same string.
We turned to Google, searching for "hot spot" and "inalámbrico". Among the 9,640 hits we found "hot spot inalámbrico" and "punto de acceso inalámbrico hot spot" (besides many pages with "inalámbrico" and "hot spot", but not in the same string). This suggested "punto de acceso" as part of the solution.
Now we searched for the exact string "hot spot inalámbrico", limiting the search to pages written in Spanish. This gave us 501 hits. "Hot spot inalámbrico" was a possible candidate, then, although not the best: it sounded too colloquial and it had far too few hits for such a widespread technology.
We tried again, still restricting the search to pages written in Spanish. This time we searched for the exact string "punto de acceso inalámbrico". Result: 155,000 hits, an excellent candidate translation.
For confirmation, we tried the translations suggested by Microsoft and by KudoZ, searching only pages in Spanish. "Zona interactiva inalámbrica" yielded 10 results in Google; "zona local de cobertura" only two hits, and both of them about cell phones, rather than wireless Internet hot spots.
So we finally chose "punto de acceso inalámbrico".
This search strategy gives excellent results and can easily be adapted to other languages and fields. An experienced translator, however, should develop a feeling for the number of hits given by a candidate term. Too few hits mean that you are on the wrong track (especially if there is a large difference between the number of hits for the source string and for the candidate term). But "too few hits" is relative: 500 hits could have been a good candidate term for some less widespread technology.