Japanese names are ordered differently than English names. The order is last name + first name, which is opposite to names in English. Japanese append suffixes to names, most commonly the honorific title "san." Automating this concatenation for names pulled from contact lists or databases generally works, but can fail in various cases. For example, a prompt that says, "Message from Tanaka," should be read "Message from Tanaka-san." There are cases where the <name> field contains "san" or a variant, so that embedding "san" into the Japanese prompt can result in duplicate titles. For example, the Japanese word for "mother" is "okaasan," so if your contacts list has the word "Mom" in it, automatically appending the "san" reads out as "Message from Okaasan-san."
If your prompt uses first and last names, don't build dependencies based on US English conventions for using first and last names. For example, If you have a prompt that says "Call <name>," Japanese are likely to say <last name><first name>(+ "san"). If they use one name, the last name is more likely for casual acquaintances and co-workers than first name.
Note: Alexa may pronounce names incorrectly or convert them to the wrong kanji character, because two names may use the same kanji pronounced differently. For example, you may say, “What is the weather in Misato?” and Alexa might respond “The weather in Sango is….” Both cities are written using the same kanji characters. You must customize your prompts to avoid incorrect pronunciation of certain characters.
Entity matching accuracy may be a challenge. Japanese liberally intersperses English phrases in movie or book titles, musician or band names, restaurant or building names, companies, etc. These words might be entered in romanized letters (romaji) or a phonetic alphabet called "katakana" (or "kana"). Unfortunately, there are multiple accepted forms of kana representation for certain phrases, so you might need to append metadata to your catalogs.
If your prompts contain numeric variables, it might generate text-to-speech (TTS) errors. Japanese appends specific suffixes to numbers, depending on the word. This structure is analogous to saying "One loaf of bread" in English. The pronunciation of the number can vary depending on the counter and even the context. For example, the characters 1日 are pronounced either "tsuitachi" when referring to the first day of the month, or "ichinichi" when referring to a duration of one day. In general, TTS should handle these issues for you. The exception is when a space is inserted in your prompts between the number variable and the counter term. In this case, the TTS likely generates an error because it tries to read the number separately from the counter.
Similarity of golden utterances in Japanese may overlap in different ways than in English. For example, in Japanese, the same verb is used for "play an artist" and "call a contact" ("Taylor Swift wo kakete" = "Play Taylor Swift," while "Taylor Swift ni kakete" = "Call Taylor Swift").
For confirmation prompts, such as "xxx, (is that) right?", you should make sure that your skill lets users reply yes or no in different ways. The literal translation "attemasuka" could let them say "attemasu" (that's right), not "hai" (yes) / "iie" (no), which causes error prompts.
It is culturally appropriate to have every error prompt in Japanese (not other languages) begin with "sumimasen" (sorry). This is the case, regardless of whether the English version has it or not. The literal translation of "I'm not quite sure how to help you with that" sounds unfriendly or odd. Although error messages are difficult because there are many cases that trigger the error prompts, they need to be proper for all of them.
In Japan, the address order starts from the postal code, then the largest unit to the smallest unit. In addition, Tokyo is technically a prefecture (equivalent to state) due to its size. As a result, prompts that use city and state information in the US might have unintended results in Japanese. For example, asking for the weather in Tokyo generates an error prompt that asks you to be more specific, in the same way that asking for the weather in Texas would.