Gemini 4 Neon vs Claude Fable 5.2 vs GPT-6.2 AstraResearching ways of communicating that come across as kind, polite, and competent in Japanese business
We tested how far Google's Gemini 4 Neon goes on the same research request. We compared it with Anthropic's Claude Fable 5.2 and OpenAI's GPT-6.2 Astra. Between 09:46 and 09:49 JST on 20 November 2026, we sent the same request, written in Japanese, to three threads. Every model ran at the "high" reasoning setting. All three used the same research skill (d-search) with the instruction to keep costs down (go cheaper).
For important research where thorough coverage matters, we still recommend using more than one model.
For important research where thorough coverage matters, we still recommend using more than one model.
Neon's output text is the easiest to read and hits the key points.
Another option is to have multiple models run the research and then have Neon edit the results into a report.
Brief · where Neon sits
Where Neon sits compared with the other two.
Neon compared with the other two
Each row is one measure. Zero is at the left end of the line, and each point's position shows the value for that measure. The blue point is Neon, and the grey points are Fable and Astra. Scales differ by row, so compare point positions only within the same row.
Work timeAbout 4 minutes shorter than Fable. About twice Astra's
0
Astra 6 min 40 s
Fable 17 min 45 s
Neon 13 min 35 s
Tokens used60% of Fable's. About 1.7 times Astra's
0
Astra 2.73M
Fable 7.64M
Neon 4.6M
Fully backed citationsSame share as Fable. Astra had all of its citations backed
0
Fable 4 of 7
Astra 5 of 5
Neon 4 of 7
Evidence in the noteAbout half of Fable's. 1.6 times Astra's
0
Astra 10
Fable 33
Neon 16
NeonFable, Astra
Neon13 min 35 s, 4.6M tokens. No usage reported. Note uncommittedFable17 min 45 s, 7.64M tokens. Asked once when it reached its budget. Also listed what it did not findAstra6 min 40 s, 2.73M tokens. Asked no questions. Found 5 cases with the receiver's own words
Research, based on real examples, what communication styles make Japanese people feel that someone is kind, polite, competent, and understanding in a Japanese business environment
$d-search go cheaper
"$d-search" calls Dorado's research skill. It quotes the original text of each source and keeps confirmed facts apart from the AI's inferences. "go cheaper" asks for cheaper research methods without changing the scope or depth.
Measured · elapsed time
Which model finished first?
13 min 35 sNeonanswered at 10:04
17 min 45 sFableanswered at 10:07
6 min 40 sAstraanswered at 9:56
Work time
Each bar is the work time from sending the request to the last answer. Time the model spent waiting for our reply is not in the bar and shows as a dotted line.
Neon
waiting 48 s
13 min 35 s. Answered at 10:04. Asked whether it could continue saving, and got our reply 48 s later
Fable
waiting 13 s
17 min 45 s. Answered at 10:07. Reached its 4M-token budget and asked how to proceed, and got our reply 13 s later
Astra
6 min 40 s. Answered at 9:56. Asked no questions
0 min5 min10 min15 min20 min25 min
Time spent waiting for our reply
Astra asked nothing and finished in 6 min 40 s. Neon took 13 min 35 s, and Fable took 17 min 45 s.
Neon asked once whether it could continue saving the research note, and we chose "continue and allow saving". Fable asked once when it reached its budget.
Measured · tokens
How many tokens did each run use?
4.6MNeonnot reported
7.64MFablereported 6.9M
2.73MAstrareported about 2.4M
Tokens in the main session and the child tasks
Usage in the logs. The dark part is the main session, and the light part is the child tasks. The black tick is the figure the model itself reported last. The dashed line is the 4M-token budget that all three set before researching.
Neon
Main 4.02M
4.6M (main 4.02M, 0.58M in 2 child tasks on Flash). Not reported
Fable
Main 3.45M
4.19M
7.64M (main 3.45M, 4.19M in 4 child tasks on Sonnet 5.5). Reported 6.9M
Astra
Main 2.33M
2.73M (main 2.33M, 0.4M in 1 child task on GPT-6.2 Astra). Reported about 2.4M
02M4M6M8M
Main sessionChild tasks (Neon: Flash. Fable: Sonnet 5.5. Astra: GPT-6.2 Astra)Budget (4M tokens for all three)The model's own report
All three set a budget of 4M tokens before they started researching. Astra used 2.73M, staying within 68% of its budget. Neon used 4.6M, 15% over budget, and Fable used 7.64M, 91% over.
Neon's answer gives no usage figure. At the end of its answer, it writes that it could not get a measured value because the meter script (meter.py) does not read Antigravity logs. Counting from the logs gives 4.02M for the main session and 0.58M across 2 child tasks on Flash, for a total of 4.6M.
Fable asked once how to proceed when it hit 4M (4.3M in the logs). Research was done and only writing remained. We chose "research the items not found too, then finish". Fable's final reported figure was 6.9M, 90% of the logged total. Astra reported about 2.4M, 88% of its logged total.
The share taken by child tasks also differed. Child tasks accounted for 55% of Fable's usage, 13% of Neon's, and 15% of Astra's.
Measured · link check of the answers
Does each chat answer hold up as written?
After the runs, we opened every cited link in the answers and compared each page with the passage that cites it. A passage counts as backed when every figure, name, date, and condition in it is on the page.
Of the 19 cited passages in the three answers, 13 were fully backed. The remaining 6 were partly backed. None were unbacked, dead links, or unverifiable.
4 of 7Neonfully backed
4 of 7Fablefully backed
5 of 5Astrafully backed
How well the cited passages are backed
For each model, the share of its cited passages that were fully backed, partly backed, or not backed or unverifiable.
Neon
4
3
4 of 7
Fable
4
3
4 of 7
Astra
5
5 of 5
0%25%50%75%100%
BackedPartly backedNot backed or unverifiable
In all 6 partly backed passages, the quoted text itself matched the page. What did not match was the heading or the note the model added under the quote.
Neon: in its criteria for a good email, it wrote that "concise key points" was "first place". The page gives only the figure 85.05% and does not state a rank. The 2025 survey trend added in its note is not on this page either.
Neon: in its note on unpleasant emails, it cited 45.85% from the 2024 survey. That figure is not on the cited page.
Neon: it presented the quote from the Sales Practices Survey as being about corporate customers as a whole. The quoted sentence is about customers who expect "deep relationship-building sales", and the "60% continuing 5 or more times" in its note is also for that group. The contrast "rather than explaining their own products" is not on the page either.
Fable: it described the evaluation of Sony Bank as the words of "judges". The page says both general consumers and judges evaluate, and does not state which group wrote this passage.
Fable: it gave the source location for the RENGO survey as "p. 1-2". The quoted passage is on page 4 of the PDF.
Fable: in its note on overly polite wording, it cited the Agency for Cultural Affairs' 2018 report. That report is not on the cited page.
Neon's 2024 figure and 2025 trend, and Fable's Agency for Cultural Affairs report, each have separate source links in their research notes. In the answers, the models wrote them under a quote from another page without giving those links, so they were graded as partly backed. This check covers only the links in the answers.
All 5 of Astra's cited passages are quotes from one individual and carry no percentage figures. Neon's and Fable's cited passages include survey figures and add more figures in their notes. The more points there are to check, the easier it is for a mismatch to appear. This is our reading.
Dates also had small discrepancies. Astra wrote dates not found on the linked page in 2 items: "April 2023" for Builnet and "1 December 2024" for the mock-email study. We did not lower the grade for dates alone.
There were 4 factual sentences without a source link in Neon, 4 in Fable, and 2 in Astra. This count is not part of the grading.
Compared · what each answer found
Where the evidence chosen by the three answers overlaps, and where it parts.
Types of evidence cited
The 19 passages cited in the answers, grouped by type of evidence. Each dot is one cited passage. For Neon and Fable, 5 of 7 were questionnaire surveys with percentage figures. For Astra, 4 of 5 were company cases with the receiver's own words.
Type of evidence
Neon
Fable
Astra
Questionnaire survey (percentage figures)
5
5
0
Research paper
2
0
1
Third-party rating or user review
0
2
0
Company case (receiver's own words)
0
0
4
Total
7
7
5
Evidence in the answers
The items of evidence that appeared in the three answers. A filled dot means the model cited that item as a source. A ring means it only mentioned the item in a note or summary.
Evidence written in the answer
Neon
Fable
Astra
Total
Email surveys (Japan Business Mail Association)
Top unpleasant content: "not answering the question" 44.29%, 2nd "overbearing phrasing" 37.3% (Business Mail Fact-Finding Survey 2026)
2
"Overly polite to the point of seeming insincere" (45.52%) is a clue that AI wrote the email (same survey)
2
79.48% do not tell the sender even when they feel unpleasant (same survey)
2
A good email has "key points summarized concisely" 85.05% (same survey)
Over 60% feel a reply is slow if it does not arrive within 24 hours (same survey)
1
Other surveys
Proactive horenso before being asked and fast responses build a manager's trust (Persol Research and Consulting, Kyushu University, 2025)
1
What corporate clients expect of sales: "presenting hypotheses for value creation" and "organizing diverse internal opinions" (Persol Research and Consulting, Sales Practices Survey 2025)
1
1st place for "a competent person" is "fast work and responses" (Biz Hits, 2021)
1
Sales reps trusted by executives: considers issues together 51.4%, fast to respond 49.5% (Softbrain Service, 2023)
1
Unforgivable complaint handling: passed around 41.8%, spoken to as if at fault 35.3% (RENGO, 2017)
1
1st place for a dependable manager is "listens to me" 32.6% (Biz Hits, 2025)
1
Research papers and government reports
Propose candidate dates yourself in a rescheduling email (Yokokawa, 2023, Journal of Japanese Language Teaching, No. 185)
1
Point out a business partner's mistake without flatly asserting it, using "my understanding is" and "I think" (Yamashita and Ono, University of Tsukuba)
1
In a mock email, "felt politeness" from a thank-you for arranging extra items (Ikeuchi, 2024)
1
Honorifics can "put up a wall between people" (Agency for Cultural Affairs, 2018 report)
1
Third-party ratings and user voices
Evaluation of Sony Bank's desk: "accurately grasping the customer's intent" and "proactive provision of information" (HDI-Japan, 2026)
1
Auto insurance user review: "handled it kindly" (Oricon, 2015)
1
Company cases (receiver's own words)
Builnet: a colleague sent the documents, and she "felt the warmth of colleagues helping one another"
1
Osaka Energy Service: on the proposal to keep a page slated for removal, "a professional's perspective really is different"
1
Tsubakimoto Kogyo: the salesperson's phrase "In short, your company is a problem-solving company" clicked right into place
1
Frontline Works: weekly meetings made clear progress, what was missing, and the next tasks
1
Cited as a sourceMentioned in a note or summary
0 items of evidence used by all three, 3 used by two (all Neon and Fable), and 19 used by only one (Fable 8, Neon 6, Astra 5).
Neon and Fable started from the same survey: the "Business Mail Fact-Finding Survey 2026" by the Japan Business Mail Association. The two answers quote the same two passages from it. One is that the top reason receivers felt unpleasant was "not answering the question" (44.29%). The other is that being "overly polite to the point of seeming insincere" (45.52%) serves as a clue that an email was written by AI. Their conclusions were close too: both pointed not to the volume of honorifics, but to answering what was asked and responding quickly.
After the email survey, the three diverged.
Neon: cited two research papers: proposing candidate dates yourself in a rescheduling email (Yokokawa, 2023, *Journal of Japanese Language Teaching*, No. 185) and pointing out a business partner's mistake without flatly asserting it, using "my understanding is" and "I think" (Yamashita and Ono, University of Tsukuba). It also used two surveys from Persol Research and Consulting.
Fable: alongside surveys, it used a third-party evaluation of a customer support desk (HDI-Japan's rating of Sony Bank), a user review (Oricon), and a consumer survey on complaint handling (RENGO). It added a "What was not found" section and stated that it did not find company cases showing a staff member's wording together with how it was rated.
Astra: used no percentage figures at all. It chose five cases where receivers described their evaluation in their own words: four named company cases (Builnet, Osaka Energy Service, Tsubakimoto Kogyo, Frontline Works) and one mock-email study. It placed an English translation under each Japanese quote.
None of the evidence in Astra's answer appears in either of the other two answers. Astra included the 85.05% "concise key points" survey result in its research note, but did not use it in its answer.
Astra did find material close to the kind of case Fable said it could not find: client interviews on web-production agencies' sites and an employee interview on a recruiting site. Even so, among the five cases, only one (Tsubakimoto Kogyo) includes the exact words the staff member actually spoke.
What follows is our reading. For someone writing business emails or running meetings in Japan, how usable each answer is split as follows.
Astra: gives phrasing examples for each situation. For example, instead of ending at "things are going smoothly", saying "The body text is finished. What remains is checking the price table. If you can confirm the amounts by Thursday, we can publish on Friday." Astra wrote these examples itself; they are not quotes. Astra notes that the examples do not represent the preferences of Japanese people as a whole.
Neon: pulled wording actually used from research papers, such as adding "if another time outside the options above works better for you" after listing candidate times, and pointing out a deadline mistake with "My understanding is that the agreed deadline was today." Because it also includes survey percentages, it works as material for explaining the point inside a company.
Fable: put the behaviors for the four impressions in a table and stated the target group and limits for each survey. It gives no phrasing examples. It fits readers who want to verify what each survey does and does not measure.
The main result · research notes
The research note each run wrote.
The answer each run returned in chat is a summary of the research note it wrote. For each item of evidence, the note holds the source's original text and where in the source it appears. You can read all three notes in full under "Actual results" below.
The number of evidence items in each run's research note, and the note's length in characters.
Neon ran its own script to check its 16 quotes against the original text, and reported finding and fixing one single-character difference. It did not commit the note. A file showing that Git was in the middle of another operation (.git/index.lock) had been left behind, and Neon noted this on the first line of its answer.
Fable reported adding one paragraph with this run's measurements and a caution to the documentation of the skill that sets research budgets. That edit was not part of the request.
Each run's research note and final answer are shown unedited. Switch between the models, and between "Research note" and "Final answer", to compare them.
Done: Finished researching communication styles that come across as "kind, polite, competent, and understanding" in Japanese business settings and saved the research note (only git commit and push were not run because .git/index.lock dated 2026-11-20 10:01:23 remained).
In conclusion, the style evaluated as "kind, polite, competent, and understanding" in Japanese business settings is determined not by the volume of honorifics, but by a combination of structure that reduces the receiver's burden of reading, thinking, and replying (concise key points, direct answers to questions, offering options and a way out, fast initial responses and proactive horenso [reporting, contacting, and consulting]) and phrasing that protects the other party's standing and face (pointing out mistakes without flatly asserting them, designing requests around the other party's convenience). Conversely, excessive politeness that feels out of place for the sender's role, or messages that fail to answer questions or lack needed information, quietly lower evaluations without being pointed out in roughly 70 to 80% of cases.
Main surveys and cases shaping the conclusion (7 items)
In the Japan Business Mail Association's 2026 survey (1,293 respondents), "concise key points" was first place at 85.05% as the criterion for a good email
"For what made a business email feel well written, 85.05% named 'key points are summarized concisely', and this evaluation standard is shared regardless of whether AI is used."
Business Mail Fact-Finding Survey 2026 | General Incorporated Association Japan Business Mail Association › Silent evaluations determine business trust · published 2026-06-01 (updated 2026-06-30) · sourceAI: This is the 2026 survey by the General Incorporated Association Japan Business Mail Association. Consistent with the 2025 survey's tendency that concise key points and clear structure matter more than etiquette and politeness, it shows that the core of appearing competent is concise organization of key points. It does not show percentages for spoken conversation.
In the Japan Business Mail Association's 2026 survey, the top unpleasant email issue was "not answering the question" (44.29%), and second was "overbearing phrasing or a condescending tone" (37.3%)
"The top item that made receivers feel unpleasant was 'not answering the question' (44.29%), and second was 'overbearing phrasing or a condescending tone' (37.3%). What stands out is that matters of communicative attitude and sincerity rank higher than technical mistakes such as typos or forgotten attachments."
Business Mail Fact-Finding Survey 2026 | General Incorporated Association Japan Business Mail Association › Failures lower your evaluation out of sight · published 2026-06-01 (updated 2026-06-30) · sourceAI: This shows that what most makes Japanese receivers feel someone is "not understanding or insincere" is not typos, but failing to answer questions (44.29%) and an overbearing tone (37.3%) (in the 2024 survey, the top unpleasant item was "lacking necessary information" at 45.85%). It does not directly measure sufficient conditions for an active "kind" rating.
In Yokokawa's (2023) study of rescheduling emails (16 native speakers and 18 non-native speakers working in Japan), proposing candidate times and "offering compensation" were practiced as consideration to reduce the other party's burden
"In the open-ended questionnaire, businesspeople who proposed candidate dates and times gave reasons such as: suggesting dates themselves cuts down the number of exchanges, avoids troubling the other party as much as possible, gives the other party the choice, and makes it easier for the other party to reply."
Consideration Language Behavior of Native and Non-Native Japanese Speakers in the Structure and Content of Japanese Business Emails (Journal of Japanese Language Teaching, No. 185) › 4-2-1 Content mentioned from codes within structural elements · published 2023-08 · sourceAI: This shows real examples of offering 2 to 3 candidate times yourself to cut back-and-forth effort, while adding an offer of compensation such as "If another time outside the options above works better for you..." to avoid being pushy. The same paper analyzes how "I am terribly sorry to trouble you, but could you please adjust your schedule to one of the dates and times below?" can prioritize one's own convenience in content even while using honorifics. Impression experiments on the reader's side are left as future work.
In Yamashita and Ono's (University of Tsukuba) utterance study (20 businesspeople in their 20s and 30s), native Japanese speakers avoided flat assertions by using "my understanding is" (ninshiki da) and "I think" (to omou) when pointing out a business partner's misunderstanding
"(6) and (7) use expressions such as 'my understanding is' and 'I think' to carry the implication that 'I only think so, and it may not actually be the case', avoiding a flat declaration of error and thereby trying to soften the threat to the other party's negative face."
On Considerate Expressions in Business Situations by Native and Non-Native Japanese Speakers › 4.2 Scene 2: Pointing out a mistake · no publication date stated (accessed 2026-11-20) · sourceAI: This shows that when correcting a business partner's mistaken deadline, native Japanese speakers did not use the textbook "-yoda" ("it seems") (0 people), and instead made polite corrections that protected the other party's face using "my understanding is" (3 people) and "I think" (2 people), such as "My understanding is that the agreed deadline was today." Sample size is 10 per group, with written responses.
In the Persol Research and Consulting and Kyushu University (2025) longitudinal study (5 organizations, 304 leaders and 1,848 members), proactive horenso before being asked and fast responses build trust, while going beyond instructions can backfire
"To earn a manager's trust, proactive behavior such as reporting, contacting, and consulting at the right time before being asked, and prompt responses, were confirmed to be effective. On the other hand, even bold efforts intended to contribute to a leader can damage trust if they go beyond the scope of instructions, so care is needed in day-to-day communication."
"Research on Trust Relationships between Managers and Subordinates" Announced - Persol Research and Consulting › For Members › Proactive behavior builds trust · published 2025-02-13 · sourceAI: This shows that timely horenso before being asked and fast responses link directly to being rated "competent and trustworthy" at work, along with the boundary condition that jumping ahead beyond agreed scope without alignment can hurt trust instead. The data comes from 5 organizations (with a high share of male leaders).
In Persol Research and Consulting's "Sales Practices Survey 2025" (3,000 respondents total, including 1,000 corporate buyers), B2B customers valued "presenting hypotheses for value creation" and "organizing diverse internal opinions" rather than explaining one's own company
"Among the behaviors expected of sales from problem formulation through the proposal stage, 'presenting hypotheses for value creation' and 'organizing diverse internal opinions' are emphasized in order to 'propose solutions from a sharp professional perspective'."
"Sales Practices Survey 2025" Announced - Persol Research and Consulting › Survey Results Summary › 1. Expectations for relationship-building behaviors that engage deeply with customers · published 2025-11-11 · sourceAI: This shows that for B2B buyers making purchases of 1 million yen or more, the "understanding and professional" rating that leads to long-term business (60% continuing 5 or more times) comes not from introducing one's own products, but from presenting hypotheses about the customer's problems and organizing opinions inside the customer's company. Low-price and standard-product transactions are excluded.
Counter-evidence: In the Japan Business Mail Association's 2026 survey, "phrasing that does not fit the person's role" (48.28%) and "overly polite to the point of seeming insincere" (45.52%) act as clues of AI authorship and unnaturalness, and 79.48% of unpleasant impressions lower evaluations without being pointed out
"44.86% of people say they have noticed 'an AI must have written this' in an email received from someone else. The clues that give it away are human mismatches such as 'phrasing that does not fit the person's character or standing' (48.28%) and 'overly polite to the point of seeming insincere' (45.52%)."
Business Mail Fact-Finding Survey 2026 | General Incorporated Association Japan Business Mail Association › Silent evaluations determine business trust · published 2026-06-01 (updated 2026-06-30) · sourceAI: This shows that overly humble honorifics and boilerplate out of step with the sender's role are seen through as unnatural rather than "polite". Because 79.48% in the same survey silently lower their evaluation without telling the sender even when they feel unpleasant, excessive honorifics and insufficient checking cause a loss of trust without the sender realizing it.
Summary by the 4 impressions (AI only)
"Competent" (sources: 1 and 5 above, plus Findings 2, 7, and 12 in the note): Concise key points and clear structure; a fast initial reply stating "by when I will answer" even when you cannot answer immediately; timely horenso before being asked (while avoiding solo actions beyond the scope of instructions).
"Kind" (sources: 3 above, plus Findings 7, 8, and 10 in the note): Proposing multiple candidate times yourself to save the other party back-and-forth effort while adding a way out (compensation) if none fit; cutting long apologies and unnecessary detail in already-apologized matters to spare the reader fatigue; noticing follow-up help that people around you could use right now.
"Polite" (sources: 3, 4, and 7 above, plus Findings 9, 11, and 15 in the note): Not only honorific forms, but content designed to leave the decision with the other party; phrasing that avoids flat assertions when pointing out a mistake, such as "my understanding is" or "I think"; honorifics that fit the role and situation without going overboard.
"Understanding" (sources: 2 and 6 above, plus Findings 3, 4, 13, 14, and 16 in the note): Answering the other party's question head-on with all needed information; offering problem hypotheses and organizing internal opinions for corporate clients rather than pitching your own company; recognizing the asymmetry that managers focus on "what you do" (horenso and execution) while subordinates focus on "how you engage" (explanation, listening, support); checking intent with options such as "A or B?" before rework when instructions are unclear.
Detailed note with all 16 quotes, English translations, and limits
The exact-match verification script detected a one-character difference in Finding 5 ("にのぼり、" -> "にのぼります。") and fixed it to exact match (16/16).
As an M3 and date check, changed the heading of Finding 7 to the survey item name itself, and removed an unlisted publication year in Finding 15, setting it to Not stated.
Tokens: actual count not obtained because meter.py does not measure Antigravity logs / budget 4M / about 12 main turns (+ searches from 2 stopped lightweight flash subagents) / about 17 min elapsed (including waiting for write-permission confirmation).
Done: Finished the research and saved a note with 33 items of evidence. I did not find company cases that include the staff member's actual wording.
What gave Japanese receivers a good impression was not someone who used a lot of honorifics. Five things came up repeatedly in the surveys: answering what was asked, responding quickly, grasping the other party's situation ahead of time, listening, and not looking down on, blaming, or brushing off the other person. Overly polite wording, on the other hand, is taken as unnatural.
Evidence that shaped the conclusion
Japan Business Mail Association's 2026 survey: the top unpleasant issue is "not answering the question"
"The top item that made receivers feel unpleasant was 'not answering the question' (44.29%), and second was 'overbearing phrasing or a condescending tone' (37.3%)."
Business Mail Fact-Finding Survey 2026 › Failures lower your evaluation out of sight · 2026-06-01 · source
AI: Answers from 1,293 people who use email at work. Answering the question and not talking down matter more than typos. In the same survey, over 60% feel a reply is slow if it does not arrive within 24 hours, and 79.48% do not tell the sender even when they feel unpleasant.
Biz Hits' 2021 survey: first place for "a competent person" is "fast work and responses"
"Someone who replies quickly to anything. Of course for things they can answer right away, and for things they don't know, they still send an initial reply saying by when they will answer (woman in her 20s, IT sales)"
Ranking of traits of competent people around you › [1st place: Fast work and responses] · 2021-10-04 · source
AI: A survey of 500 working people; the quote is one respondent's answer under first place. It shows the behavior of sending back a deadline for the answer even before the answer is ready. The survey is about workplace colleagues, not evaluations of business partners.
HDI-Japan's judges evaluate Sony Bank's support desk (2026)
"The handling is clear and smooth, accurately grasping the customer's intent and closing the psychological distance. Explanations backed by deep knowledge and proposal skills, along with proactive provision of information, also create a sense of trust."
HDI's evaluation of Sony Bank › Inquiry desk · 2026 rating · source
AI: Evaluation by judges who actually made inquiries, published in Sony Bank's press release. Grasping intent and providing information proactively are cited as reasons for trust. The staff member's own words are not recorded.
User review of Mitsui Sumitomo Insurance (Oricon, 2015)
"It was an accident right after I signed up and I was completely at fault, so I was glad they handled it kindly. (Age 60+ / female)"
Call center handling satisfaction ranking and reviews › Mitsui Sumitomo Insurance · 2015 edition · source
AI: "Kindly" is the user's own word. That not blaming a person who is at fault was received as kindness is my reading. This is one person's comment selected by Oricon.
Softbrain Service's 2023 survey: sales reps trusted by executives
"When asked 'Q4. Please tell us the characteristics of business-partner sales reps you feel you can trust (multiple answers)' (n=105), 51.4% answered 'considers and proposes ways to solve our company's issues together', 49.5% answered 'fast to respond by phone or email', and 45.7% answered 'explains not only the advantages but also the disadvantages'."
Characteristics of trustworthy sales representatives · 2023-07-24 · source
AI: Answers from 105 executives of BtoB small and medium-sized enterprises. First place for untrustworthy reps was "makes excuses or brushes things off" at 56.2%. What was measured is "trust", not words such as "kind".
"When general consumers (1,000 people) were asked what kind of handling by store clerks or staff they found unforgivable in response to complaints, 'being passed around from one department to another' (41.8%) was the most common, followed by 'being spoken to as if I were at fault' (35.3%), 'being addressed in rude language' (34.4%), 'only apologizing and taking no other action' (32.3%), and 'not getting a reply for a long time' (27.7%)."
Fact-Finding Survey on Consumer Behavior › p. 1-2 · 2017-12-21 · source
AI: Being passed around and being spoken to as if at fault rank above rude wording. Just apologizing is not enough. This covers complaint handling at stores and service desks, not business-to-business dealings.
Japan Business Mail Association's 2026 survey: overly polite wording
"The clues that give it away are human mismatches such as 'phrasing that does not fit the person's character or standing' (48.28%) and 'overly polite to the point of seeming insincere' (45.52%)."
Business Mail Fact-Finding Survey 2026 › Silent evaluations determine business trust · 2026-06-01 · source
AI: These are shares among people who have noticed that an email was written by AI (44.86%). The question did not directly ask whether people like or dislike overly polite wording. The Agency for Cultural Affairs' 2018 report also notes that honorifics can "put up a wall between people".
AI only (my summary)
Impression
Behavior that gave the receiver that impression
Evidence
Kind
Do not blame even when the other person is at fault. Offer alternatives. Explain disadvantages too.
4, 5
Polite
Start by checking the situation. Explain clearly until the other person is satisfied. Not the volume of honorifics.
3, 7
Competent
Reply quickly. When the answer is not ready, send back a deadline. Keep key points concise.
1, 2, 5
Understanding
Grasp the intent behind the question. Listen to opinions before answering. Think through problems together.
1, 3, 5
The note also includes a result where first place for a "dependable manager" is "listens to me" at 32.6% (Biz Hits, 2025).
Most surveys measure "discomfort", "trust", or "satisfaction". The only sources that use the four target words themselves are the HDI evaluation, the Oricon review, and the Biz Hits survey. Many of the survey sponsors are companies selling training or sales support, so the percentages are rough guides.
What was not found
Company cases that present a staff member's wording together with how it was evaluated (such as JAL, ANA, or the Imperial Hotel).
Japanese studies measuring listener evaluations of cushion words, repeating back, backchanneling, or apology phrasing. The only one with numbers was a single experiment on nodding.
Surveys comparing the impressions of "ryokai desu" ("got it") and "shochi shimashita" ("understood, certainly").
The same request is also running in Japanese business communication research (Gemini) and Japanese Business Communication Styles (Codex). I treated it as a model comparison, proceeded, and saved the note under a separate name. I added one paragraph to the research-token-budget skill with this run's measurements and a caution that the fetch script overwrites saved pages in the same folder.
Actual token usage: 6.9M (73% above plan)
Plan: 4M
What receivers in Japanese workplaces and business relationships praised were situations where kind meant "they even thought of the trouble it would take me", polite meant "they noticed what I had done", competent meant "they gave me reasons that helped me decide", and understanding meant "they put into words what I could not explain well myself". The following five examples show these differences in the receivers' own words.
Feels kind: take on a concrete piece of effort tailored to the other person's circumstances.Mayumi Makino of Builnet had her husband fall ill while she was on maternity leave, and needed documents for paperwork. A colleague sent the documents so Makino would not have to come in to the office, and called her as well. Looking back, she says, "I felt the warmth of colleagues helping one another, rather than a merely administrative relationship." AI reading Go one step beyond "let me know if you need anything" and offer "I can send the documents from here." It also saves the other person the effort of having to ask for help again. Builnet employee interview with Ms. Makino, "The warmth of colleagues helping one another eased my anxiety about returning to work after maternity leave", April 2023
Feels polite: state specifically what you are thanking them for.In an email study simulating a product shortage, an experienced practitioner on the side that arranged the extra shipment answered, "It was our mistake, of course, but because they thanked us for being able to send 5 extra packs, I felt politeness." The same person took a thank-you for a routine check as leaving a plain impression. AI reading If the other person coordinated shipping inside their company, write "Thank you for arranging the shipment." That shows you also recognize what they did for you. This is not an actual delivery incident, but the evaluation of a person who received a mock email. Yuka Ikeuchi, "Actual State of Communication in Business Emails in Complaint Situations", Section 4.2.3, p. 155, 1 December 2024
Feels competent: even when proposing something different from the request, give reasons the other party can use to decide.Osaka Energy Service intended to cut its explanatory page on district heating and cooling, thinking specialist business partners already knew it. The web production agency showed that general readers also read it and that the page ranked high in search and served as an entry point to the site. The company kept the page, and Matsuhisa of the Technical Planning Department praised the series of proposals, saying, "I was impressed that a professional's perspective really is different." AI reading Expertise came across when they showed "who comes and from where", something the client had not factored into its judgment. The client can then explain the reason for adopting the proposal themselves. Osaka Energy Service representative interview, "Selecting content through site analysis", accessed 20 November 2026
Feels understood: sum up the other person's concrete examples in one phrase that clicks with them.A representative at Tsubakimoto Kogyo explained how each department handled customers' problems, yet struggled to express the company's overall appeal. "Then, at the end, the salesperson summed it up in one phrase: 'In short, your company is a problem-solving company.' It clicked right into place for me." AI reading That single phrase resonated because the work of the multiple departments the representative described shared a common thread. In the website production that followed, that idea was also used to decide what to put on the site. Tsubakimoto Kogyo representative interview, "What did you prioritize in production?", accessed 20 November 2026
Feels polite and dependable: in progress updates, state how far things have come, what is missing, and what comes next.Nagafuchi of Frontline Works said of GIG, which handled their website production, "Not only did they keep in frequent touch, but in our weekly meetings they always made clear 'how far things have progressed', 'what is missing', and 'what we will do next'." The evaluation was that they could leave the work in GIG's hands with peace of mind, without worrying about the status through to the end. AI reading For example, instead of ending at "things are going smoothly", say "The body text is finished. What remains is checking the price table. If you can confirm the amounts by Thursday, we can publish on Friday." The other party then knows what they need to do and by when. GIG, Frontline Works case-study interview, answer on progress management, 3 April 2026
To apply these differences in practice, the main thing to review is the sentence right after "thank you", "understood", or "things are going smoothly". What the other person did for you, how you understood it, or who will do what next: making one of those concrete to fit the moment conveys the substance of your care and understanding. This is an interpretation from cases told by Japanese receivers, and does not show the preferences of Japanese people as a whole or differences from other countries.
同じ依頼は 日本のビジネスコミュニケーション調査 (Gemini)と Japanese Business Communication Styles (Codex)でも動いています。モデル比較と判断して進め、ノートは別名にしました。research-token-budget スキルには、今回の計測値と、取得スクリプトが同じフォルダの保存ページを上書きする注意を1段落足しました。
親切に感じる。相手の事情に合わせて、具体的な手間を引き取る。ビルネットの牧野真弓さんは、産休中に夫が病気になり、手続きに書類が必要になりました。同僚は牧野さんが会社に出向かずに済むよう書類を送り、電話もくれました。本人は「事務的なやりとりをする間柄ではなく、仲間として助け合う温かさを感じました。」と振り返っています。 English translation: “I felt the warmth of colleagues helping one another beyond routine administrative exchanges.” AIの読み取り 「何かあれば言ってください」より一歩踏み込んで、「書類はこちらから送れます」と申し出る。相手は改めて助けを頼む手間も省けます。ビルネット・牧野さんの社員インタビュー「仲間として助け合う温かさが、産休後の職場復帰に対する不安を解消してくれた」、2023年4月
丁寧に感じる。何に対して感謝しているのかを言う。商品が不足した場面を想定したメール研究で、追加分を手配した側の実務経験者は「こちらのミスではあるんですけど、5 パック追加で送れることに対してのお礼があったので、丁寧さを感じた。」と答えました。同じ人が、一般的な確認へのお礼はあっさりした印象だと受け取っています。 English translation: “It was our mistake, but they thanked us for being able to send the five extra packs, and I felt that was courteous.” AIの読み取り 相手が社内で出荷を調整してくれたなら、「出荷を調整してくださり、ありがとうございました」と書く。相手が何をしてくれたか分かっていることまで伝わります。これは実際の納品事故ではなく、模擬メールを受け取った本人の評価です。池内優香「クレーム場面におけるビジネスメールでのコミュニケーションの実態」4.2.3節、155頁、2024年12月1日
仕事ができると感じる。依頼と違う案でも、相手が判断できる理由を添える。大阪エネルギーサービスは、専門の取引先なら既に知っているという理由で、地域熱供給の解説ページを削るつもりでした。制作会社は、一般の人も読むこと、そのページが検索上位に出ていてサイトの入口になっていることを示しました。同社はページを残し、技術企画部の松久さんは一連の提案を「やはりプロの目線は違うなと感心しました。」と評しています。 English translation: “I was impressed by how differently a professional sees things.” AIの読み取り 専門性が伝わったのは、発注者が判断に入れていなかった「誰が、どこから来るか」を示した場面です。相手は提案を採る理由を自分でも説明できます。大阪エネルギーサービス担当者インタビュー「サイト分析によるコンテンツの取捨選択」、2026年11月20日閲覧
理解してくれたと感じる。相手が話した具体例を、本人が納得できる一言にまとめる。椿本興業の担当者は、各部署が顧客の悩みにどう対応しているかを説明したものの、会社全体の魅力をうまく言い表せずにいました。「すると、営業さんが最後に『つまり、御社は課題解決企業なんですね』と一言でまとめてくれたんです。ストンと腹に落ちましたね。」 English translation: “At the end, the salesperson summed it up in a few words: ‘So your company solves problems for its customers.’ That really clicked for me.” AIの読み取り この一言が響いたのは、担当者が話した複数の部署の仕事に共通点があったからです。その後のサイト制作でも、この考え方が何を載せるかの判断に使われました。椿本興業担当者インタビュー「制作にあたって重視した点は何ですか?」、2026年11月20日閲覧
丁寧で任せられると感じる。途中報告で、進み具合・不足・次の作業を伝える。フロントラインワークスの永渕さんは、サイト制作を任せたGIGについて「こまめなご連絡だけでなく、週に一度のミーティングで『今どこまで進んでいるか』『何が足りないか』『次は何をするか』を常に明確に示してくれました。」と話しています。最後まで状況を心配せず、安心して任せられたという評価でした。 English translation: “They kept in touch and used our weekly meetings to make clear how far things had progressed, what was missing, and what would happen next.” AIの読み取り 例えば「順調です」で終えず、「本文は完成しました。残りは価格表の確認です。木曜までに金額をご確認いただければ、金曜に公開できます」と伝える。相手には、自分がいつ何をすればよいかまで分かります。GIG・フロントラインワークス事例インタビュー、進捗管理についての回答、2026年4月3日
同一份請求也在 日本商務溝通調查 (Gemini)與 Japanese Business Communication Styles (Codex)執行中。我判斷這是模型比較並繼續進行,筆記另存不同檔名。另在 research-token-budget 技能中加上了1段關於這次實測數值,以及擷取腳本會覆寫同一資料夾內已儲存頁面的注意事項。
讓人覺得親切:配合對方處境,主動分擔具體的麻煩。Builnet的牧野真弓在產假期間先生生病,辦理手續需要文件。同事為了讓牧野不必特地跑一趟公司,主動寄送文件,也打了電話關心。她本人回顧道:「不是公事公辦的關係,而是感受到了身為夥伴互相扶持的溫暖。」 English translation: “I felt the warmth of colleagues helping one another beyond routine administrative exchanges.” AI解讀 比起「有什麼需要請說」,更進一步主動提議「文件可以由我們這邊寄過去」,也能省去對方再次開口求助的負擔。Builnet牧野員工訪談「身為夥伴互相扶持的溫暖,消除了產假後重返職場的不安」,2023年4月
讓人覺得有禮周到:說清楚是針對哪件事表達感謝。在假設商品短缺情境的郵件研究中,負責安排補寄的實務經驗者回答:「雖然是我們這邊的失誤,但對方針對能補寄5包這件事表達了感謝,讓我覺得很有禮周到。」同一個人則認為,對一般確認所說的謝謝給人較平淡的印象。 English translation: “It was our mistake, but they thanked us for being able to send the five extra packs, and I felt that was courteous.” AI解讀 如果對方在公司內部幫忙協調出貨,就寫上「感謝您協助調整出貨」。這樣一來,對方也能感受到你明白他付出了什麼。這並非實際的出貨事故,而是收到模擬郵件的受訪者本人之評價。池內優香「客訴情境下商務電子郵件之溝通實態」第4.2.3節,第155頁,2024年12月1日
讓人覺得工作能幹:即使提出與委託不同的方案,也附上對方能用來判斷的理由。大阪能源服務原本認為專業往來廠商早就知道區域冷暖氣供給系統,打算刪掉解說頁面。網站製作公司則指出,一般民眾也會閱讀,且該頁面在搜尋排名靠前,是網站的入口。該公司決定保留頁面,技術企劃部的松久評價這一系列提案說:「果然專業人士的眼光就是不一樣,令我佩服。」 English translation: “I was impressed by how differently a professional sees things.” AI解讀 專業度之所以能傳達出去,在於指出了委託方原本沒納入判斷的「誰會來看、從哪裡進來」。對方自己也能向內部說明採用提案的理由。大阪能源服務窗口訪談「透過網站分析取捨內容」,2026年11月20日瀏覽
讓人覺得對方理解自己:把對方談的具體例子,歸納成讓本人心服口服的一句話。椿本興業的窗口雖然說明了各部門如何解決顧客的煩惱,卻一直難以道出整間公司的魅力。「結果業務人員最後用一句話總結:『也就是說,貴公司是一家解決問題的企業呢。』我當下就豁然開朗了。」 English translation: “At the end, the salesperson summed it up in a few words: ‘So your company solves problems for its customers.’ That really clicked for me.” AI解讀 這句話之所以打動人心,是因為窗口談到的多個部門工作確實有共通點。之後在網站製作時,這個想法也成了決定刊載內容的判斷依據。椿本興業窗口訪談「製作時重視哪些地方?」,2026年11月20日瀏覽
讓人覺得有禮周到又值得託付:期中回報時,清楚說明進度、不足之處與下一步工作。Frontline Works的永淵談到委託製作網站的GIG時表示:「不只經常聯繫,在每週一次的會議上也總是明確告訴我們『目前進行到哪裡』『缺少什麼』『接下來要做什麼』。」評價是直到最後都不必擔心狀況,可以放心託付。 English translation: “They kept in touch and used our weekly meetings to make clear how far things had progressed, what was missing, and what would happen next.” AI解讀 例如不只停在「目前一切順利」,而是說明「內文已完成,剩下價目表的確認。若您能在週四前確認金額,週五就能上線。」對方就能清楚知道自己何時該做什麼。GIG、Frontline Works案例訪談,關於進度管理的回答,2026年4月3日
Translation: For what made a business email feel well-written, 85.05% cited "the main points are summarized concisely," and this evaluation criterion is shared regardless of whether AI is used.
Translation: Furthermore, emails evaluated as "well-written" share common traits, with technical elements such as "conciseness of main points" and "clarity of structure" tending to be valued more than courtesy or politeness.
Translation: The top item recipients found unpleasant was "not answering the question" (44.29%), followed in second place by "overbearing expressions or condescending wording" (37.3%). Notably, issues related to communication attitude and sincerity ranked higher than technical mistakes such as typos or forgotten attachments.
Translation: 44.86% of respondents have noticed that an email received from someone else "was written by AI," and the clues for spotting it are human mismatches such as "wording that does not fit the sender's character or position" (48.28%) and "being polite to the point of being overly obsequious" (45.52%).
Translation: Even when they find another person's mistake, 70.4% do not point it out. Even when they felt unpleasant after receiving a work email in the past year, 79.48% did not point it out to the sender. Where does that feeling go? The answer is that it is processed in the form of "lowering one's evaluation of the other person internally."
Translation: First place for "characteristics of people who are good at their job" was "fast work and fast response." Closely following in second and third place were "high communication ability" and "able to be considerate of those around them."
Quote: 候補日時を提示した BP は,自由記述アンケートにおいて,自分から日程を提示すればやり取りの回数を減らせる・相手に極力面倒をかけずに済む・相手に選択権を与えられる・相手が返信しやすくなる,といった理由を示していた。
Translation: Businesspersons who proposed candidate dates and times stated in the open-ended questionnaire reasons such as reducing the number of exchanges by proposing dates themselves, minimizing trouble for the other party, giving the other party the right to choose, and making it easier for the other party to reply.
Translation: Therefore, while it is polite in linguistic form, in terms of content it asks the recipient to adjust to the writer's own schedule. Because the other party also has work schedules in business settings, caution is needed since such a request may be taken as the writer prioritizing their own convenience.
Translation: In the interview as well, she stated that she intentionally did not mention them because she thought a long apology including details of the internal trouble and causes of the cancellation might tire the reader before reaching the main topic of the Japanese business email (rescheduling the meeting).
Translation: Examples (6) and (7), by using expressions such as "my understanding is" (ninshiki da) and "I think" (to omou), convey the nuance that this is only what the speaker thinks and reality might differ, avoiding asserting outright that the other party is wrong and thereby mitigating threats to the other party's negative face.
Translation: To gain trust from a supervisor, proactive behavior such as reporting, contacting, and consulting at an appropriate time before being asked, along with rapid responses, was confirmed to be effective. On the other hand, even bold initiatives intended to contribute to the leader may damage trust if they go beyond the scope of instructions, so regular communication requires care.
Translation: The criteria for judging trust differ between leaders and members: leaders emphasized "what the member does," such as reporting, consulting, job execution, and responsible action. Meanwhile, members emphasized "how the leader engages," such as explanation, attentive listening, and support, confirming an asymmetry in trust criteria.
Translation: In the problem-framing to proposal stages, the behaviors expected of sales representatives for "proposing solutions from a sharp professional perspective" emphasize "presenting hypotheses for value creation" and "organizing diverse internal opinions."
Translation: Regarding honorifics in commercial fields, so that they do not become extremely excessive, it is desirable for the parties involved to consider honorific usage appropriate to interpersonal relationships and situations in that field.
Translation: In addition, quite a few respondents experienced communication gaps, such as being told when going to report to a Japanese supervisor, "Why didn't you do it that way? Please redo it," or "From a Japanese way of thinking, you should do it this way."
English translation: At the end, the salesperson summed it up in a few words: “So your company solves problems for its customers.” That really clicked for me.
English translation: They kept in touch and used our weekly meetings to make clear how far things had progressed, what was missing, and what would happen next.
English translation: When they weren’t sure what I meant by a question, they asked a follow-up. Where there was uncertainty, they checked with me and included a suggestion.
English translation: Many times, even after we thought something couldn’t be done, the production team came back a little later and told us, “We can do that machining now.”
English translation: I was a little anxious just before going back, too. When I asked Furuta, “What should I do?” she said, “You can take it easy.” Hearing that really put me at ease.
English translation: Asked what made a business email seem well written, 85.05% cited a concise summary of the main points, a criterion shared regardless of AI use.