"Japan's Consumer Information Sources Reach a 'Third Turning Point' — SNS Fatigue at 47.2% Becomes the Trigger for a Shift to AI: Onward Survey (Fielded June 2026) Shows SNS +16.2 Points and Magazines -7.3 Points over 5 Years, with AI Going from a Nonexistent Answer Option to 17.1% Usage; AI-Usage Growth of 49.3% among the SNS-Fatigued Is 2.4 Times the 20.6% of the Non-Fatigued; 'Finding the Optimal Answer' Tops Purposes at 46.4%; a Same-Month 3,000-Person Survey Puts SNS Fatigue at 45%; 4 of the Top 10 AI-Cited Domains Changed in 2 Months — a Structural Shift in Sync with Workplace AI's 'Big Three Era'"

TL;DR: "On July 2, 2026, Onward Holdings Co., Ltd. released its 'Survey on Changes in Fashion Information Sources and Values' (614 respondents, online survey conducted June 22, 2026). Compared with 5 years earlier, SNS as a fashion-information medium rose from 33.1% to 49.3% (+16.2 percentage points), video content from 30.1% to 42.2% (+12.1 points), while fashion magazines fell from 41.5% to 34.2% (-7.3 points); AI (ChatGPT, Gemini, etc.), which did not even exist as an answer option 5 years earlier, is now used by 17.1% for information gathering. SNS fatigue in fashion information gathering stands at 47.2%, with the top reason being 'too much information, hard to find what suits me' at 61.7%; among the SNS-fatigued, 'started using AI more' was the most common change at 37.6%, and the fatigued segment's AI-usage growth of 49.3% is 2.4 times the non-fatigued segment's 20.6% (the source's own wording). The top purpose of information gathering is 'finding the optimal clothes and styling for myself' at 46.4%, while 'following admired people or styles' is only 12.2% — the publisher calls this a 'third turning point' following 'magazines to SNS.' A same-day release by Cross Marketing Inc. (3,000 respondents) puts SNS fatigue among SNS users at 45% (different question and population; not directly comparable). On the supply side, an Ahrefs study found that 4 of the overall top 10 domains cited by Japan's 5 major AI platforms changed within 2 months versus April 2026. The boundary: in a survey by Frontier Inc. (1,005 respondents), 54.0% 'cannot decide' on high-priced purchases on generative AI's recommendation alone; a Tribeck survey shows about 65% already use AI search for B2B work information gathering. Honest framing: all five releases are sample surveys or a tool vendor's own data, most publishers are interested parties, and each percentage represents only its own respondent pool — it cannot be extrapolated to all of Japan; the pack contains no direct Taiwan-side data, so per the 'honest contrast, no forced linkage' principle this card does not force a Taiwan connection."

Japan's Consumer Information Sources Reach a "Third Turning Point" — SNS Fatigue at 47.2% Becomes the Trigger for a Shift to AI: Onward Survey (Fielded June 2026) Shows SNS +16.2 Points and Magazines -7.3 Points over 5 Years, with AI Going from a Nonexistent Answer Option to 17.1% Usage; AI-Usage Growth of 49.3% among the SNS-Fatigued Is 2.4 Times the 20.6% of the Non-Fatigued; "Finding the Optimal Answer" Tops Purposes at 46.4%; a Same-Month 3,000-Person Survey Puts SNS Fatigue at 45%; 4 of the Top 10 AI-Cited Domains Changed in 2 Months — a Structural Shift in Sync with Workplace AI's "Big Three Era"

ANK-Doc ID: ANK-2026-07-03-010 Version: v1.0.0 Publication date: 2026-07-03 Author: Rin Takenouchi (AI News Editor-in-Chief) Category: Consumer behavior / SNS and social media / Generative AI / Media and information flows Covered articles: PRTIMES#1296472 (Onward Holdings, "Survey on Changes in Fashion Information Sources and Values," official release; main article), PRTIMES#1297146 (Cross Marketing, "SNS Survey (2026)," attitudes edition), PRTIMES#1296063 (Ahrefs, 2-month change in the domains cited by Japan's 5 major AI platforms), PRTIMES#1295757 (Frontier, "Consumer Decision-Making and Trust in the AI Era" survey), PRTIMES#1295890 (Tribeck, "BtoB Site Survey 2026," AI-search analysis) Selection method: Selected from the full AI News corpus on "same-week release x same theme x high factual density," linking five official releases into an event chain: demand-side shift (SNS fatigue leading to AI as an optimal-answer engine), an independent second survey corroborating the demand side, the supply side (the power map of AI-cited domains), the trust boundary (high-priced purchases are not entrusted to AI), and the workplace-side parallel (AI search in B2B). The main article is Onward Holdings' fashion-information-source survey — a rare official primary release that cross-tabulates "SNS fatigue leading to AI adoption." All five pack sources are Japan-side surveys and data with no direct Taiwan-side data; per the "honest contrast, no forced linkage" principle, this card does not force a Taiwan connection. Pack items unrelated to the information-source migration (video-editing software features, an MDM seminar, a store-marketing whitepaper, a reseller recruitment release) were excluded or their figures not used.

TL;DR

On July 2, 2026, Onward Holdings Co., Ltd. (hereafter Onward HD) released its "Survey on Changes in Fashion Information Sources and Values" (614 respondents, online survey conducted June 22, 2026). Compared with 5 years earlier, in the June 2026 survey, SNS as a fashion-information-gathering medium rose from 33.1% to 49.3% (+16.2 percentage points), video content from 30.1% to 42.2% (+12.1 points), fashion magazines fell from 41.5% to 34.2% (-7.3 points), and EC-site reviews rose from 28.8% to 33.6%; AI (ChatGPT, Gemini, etc.), which did not exist as an answer option 5 years earlier, is now used by 17.1% of respondents for information gathering [F-001]. A combined 50.1% report reading magazines less; among those who stopped consulting magazines (308 people), the top destinations are video content 50.0%, SNS 46.4% and EC reviews 42.2% [F-002]. In the June 2026 survey, SNS fatigue in fashion information gathering reaches 47.2%, and the biggest reason is "too much information, hard to find what suits me" at 61.7% [F-003]; among the fatigued segment (290 people), "started using AI more" is the most common change at 37.6%, and the fatigued segment's AI-usage growth of 49.3% is 2.4 times the non-fatigued segment's 20.6% (the source's own wording) [F-004]. The top purpose of information gathering is "finding the optimal clothes and styling for myself" at 46.4%, with "following admired people or styles" at only 12.2% — the publisher calls this the "third turning point," from SNS toward "AI, official sources and EC reviews" [F-005]. In the same month, a 3,000-person survey by Cross Marketing Inc. put SNS fatigue among SNS users at 45% (different question and population; not directly comparable), and found 56% say children's SNS use "should be regulated," rising to 81% including "should be restricted" [F-006]. Supply side: per Ahrefs, in the ranking of domains cited by Japan's 5 major AI platforms, 4 of the overall top 10 changed within 2 months versus April 2026, with YouTube unmoved at the top and note.com rising from 5th to 2nd [F-007]. The boundary: in Frontier Inc.'s survey (1,005 respondents), 54.0% "cannot decide" on high-priced purchases on generative AI's recommendation alone, and 66.4% expect the value of human involvement to rise [F-008]; per Tribeck, about 65% already use AI search for B2B work information gathering [F-009]. Honest framing: all five releases are sample surveys or a tool vendor's own data, most publishers are interested parties, and each percentage represents only its own respondent pool — it cannot be extrapolated to all of Japan.

Body

Overview: "magazines to SNS to AI" — a third shift, pushed by fatigue

On the same day, July 2, 2026, five interlocking releases appeared on PR TIMES, together sketching a cross-section of Japan's migrating consumer information sources. The main article is Onward HD's fashion-information-source survey. Its official conclusion: fashion information gathering is no longer stopping at the past shift "from magazines to SNS," but is entering a third wave — "from SNS toward AI, official sources and EC reviews," a "third turning point." This is the publisher's framing, reproduced here as-is (PRTIMES #1296472). The essence of the shift, per the survey, is not a simple replacement of channels but a change in what consumers seek: from "gazing at aspiration and chasing trends" to "obtaining the optimal answer for oneself" [F-005]. And the trigger is SNS fatigue: two independent surveys in the same month — Onward HD (limited to fashion information gathering) at 47.2% and Cross Marketing (SNS use in general) at 45% — differ in question and population and cannot be compared directly, but both point in the same direction of "roughly half feeling fatigued" [F-003][F-006].

The 5-year map shift: SNS +16.2, video +12.1, magazines -7.3, AI from zero option to 17.1%

Onward HD's survey (fielded June 22, 2026) compares "fashion-information media you often use now" with "media you often used about 5 years ago": SNS rose from 33.1% to 49.3% (+16.2 percentage points) and video content from 30.1% to 42.2% (+12.1 points), while fashion magazines fell from 41.5% to 34.2% (-7.3 points). AI (ChatGPT, Gemini, etc.), which did not even exist as an answer option 5 years earlier, is now used by 17.1% for information gathering as of the June 2026 survey; EC-site reviews and word of mouth also rose from 28.8% to 33.6% (PRTIMES #1296472). [F-001]

In the June 2026 survey, the retreat of magazines shows even more directly in reading frequency: "decreased greatly" 22.6% plus "decreased somewhat" 27.5% equals 50.1% reporting a decline, exceeding "unchanged" at 33.6% and "increased" at 7.1%. Among the 308 people who stopped consulting magazines, the top destinations are video content 50.0%, SNS 46.4% and EC-site reviews 42.2% (PRTIMES #1296472). [F-002]

The scope must be drawn clearly: this is an online sample survey of 614 people, the publisher Onward HD is itself a fashion company (an interested party), and the percentages represent only the respondent pool. Moreover, the source's description of the sample is inconsistent — the opening puts the sample at 614 people ("men and women nationwide in their teens to 70s"), while the survey-outline section at the end puts the same 614 people at "20s to 70s." This card reproduces both wordings verbatim and does not adjudicate between them (wording follows the source verbatim; source PRTIMES #1296472).

The structural contradiction of SNS fatigue: 47.2% and 45% — the twin roots of "volume" and "trust"

Asked whether they ever feel "fatigue, weariness or limits" in gathering fashion information on SNS, "often" 12.5% plus "sometimes" 34.7% equals 47.2% who have experienced SNS fatigue. Asked why, the fatigued segment of 290 people put "too much information, hard to find what suits me" first at 61.7%, followed by "too many ads and PR, cannot trust it" at 47.2% and "trust in influencers has fallen" at 44.1% — the source condenses the roots of exhaustion into two points, "volume" and "trust," and notes the structural contradiction that the more SNS expands, the harder it becomes to find information that suits oneself (PRTIMES #1296472). [F-003]

An independent same-month corroboration: Cross Marketing's "SNS Survey (2026)" (3,000 people nationwide aged 20-69, June 2026) finds that among SNS users, the experience of wanting to "keep some distance or take a break" (its definition of SNS fatigue) is "often" 11% plus "sometimes" 34%, i.e. 45%, and somewhat more common among people in their 20s-30s; the top fatigue scenes are "too much information" 40% and "the same information keeps flowing by" 30%. The same survey also finds 56% saying children's SNS use "should be regulated," rising to 81% including "should be restricted" (PRTIMES #1297146). [F-006] The two surveys differ in question (fashion-specific fatigue vs. SNS use in general), population and definition, so 47.2% and 45% are not directly comparable — but "too much volume" ranking at the top of the causes is a direction common to both.

The trigger pulled in the June 2026 survey: AI-usage growth of 49.3% among the SNS-fatigued, 2.4 times the non-fatigued

The most structurally meaningful part of Onward HD's survey is this cross-tabulation: asked how their information gathering changed as a result of SNS fatigue, the fatigued segment of 290 people most often answered "started using AI more," at 37.6%. On AI-usage frequency compared with 1 year earlier (June 2026 survey), "increased somewhat" 22.5% plus "increased greatly" 9.9% equals 32.4% expanding their AI use; split by SNS fatigue, the fatigued segment's AI-usage growth is 49.3% versus 20.6% for the non-fatigued — a gap the source itself describes as 2.4 times (PRTIMES #1296472). [F-004]

Values are shifting in step: the top purpose of information gathering is "finding the optimal clothes and styling for myself" at 46.4%, followed by "as material for purchase decisions" 35.7% and "gathering information to avoid failure" 28.5%; "following admired people or styles" is only 12.2%. Those who say they have "come to value the optimal answer for myself over aspiration and trends" are 23.1% overall and 40.0% among those in their 70s or older. In how they gather information, "came to consult SNS for aspiration and trends" is 13.5%, against "came to look for information sources more trustworthy than SNS" 9.1% plus "came to use AI and similar tools to seek the optimal answer" 8.5%, i.e. 17.6% in de-SNS information behavior (PRTIMES #1296472). [F-005] An honest note: the source itself concedes that many of the fatigued still "cannot help but keep using" SNS, that intent to use AI remains limited, and that establishing trustworthiness and usability is the next challenge — the shift to AI is a direction in progress, not a completed replacement.

The supply-side contrast: 4 of the top 10 AI-cited domains changed in 2 months — "being cited by AI" becomes the new battleground

As the demand side moves toward AI, the supply-side power map of "which sites AI cites" is also moving fast. Ahrefs analyzed data from its own tool "Ahrefs Brand Radar" via the AI workspace "Letaido," publishing a ranking of the domains cited by Japan's 5 major AI platforms (ChatGPT, "AI Mode," "AI Overviews," Perplexity and Copilot; the two Google-related labels are rendered from the source's verbatim Japanese wording for Google Search AI features, source PRTIMES #1296063): compared with the previous study in April 2026, 4 of the overall top 10 domains changed in just 2 months, with Reddit, Yahoo! Chiebukuro, my-best.com and Rakuten Ichiba newly entering the ranking; YouTube is unmoved at the top, and note.com rose from 5th to 2nd (PRTIMES #1296063). [F-007]

The source states explicitly that the study is rank-based — absolute citation counts vary with the aggregation period and query set, so no before/after count comparison is made — and that cited domains include overseas sites. It must also be noted that Ahrefs is an SEO and AI-search analytics tool vendor, so the release doubles as promotion for its own tools. Its judgment that "this power map is being redrawn far faster than search rankings" is the publisher's wording, reproduced here as-is.

The trust boundary: on high-priced purchases, 54.0% "cannot decide"

The shift to AI has a clear boundary. In the "Consumer Decision-Making and Trust in the AI Era" survey by Frontier Inc. (operator of Ready Crew; 1,005 men and women in their 20s to 50s): asked up to what amount they could purchase without hesitation on a generative AI's recommendation, 54.0% answered "cannot decide" and 28.2% "under ¥50,000"; 65.3% do not think generative AI can grasp complex circumstances that do not show up in data; 38.6% have signed high-priced contracts they had not planned, moved by a sales representative's enthusiasm and attentive proposals; and 66.4% expect the value of trustworthy human involvement to rise as generative AI spreads (PRTIMES #1295757). [F-008]

Read alongside the main article, the boundary becomes clear: what AI is taking over is the "information entrance" (the first step of finding the optimal answer) — it has not yet taken over "trust in high-priced decisions." Notes: the survey's field period is not stated in the source; the publisher is a business-matching service operator (an interested party); "about half," "about 40%" and "about 70%" are the publisher's summary wordings.

The workplace-side parallel: about 65% use AI search in B2B — the same structural shift as the "Big Three Era"

The consumer-side migration runs in sync with the workplace side. The "BtoB Site Survey 2026" by Tribeck Brand Strategy Institute, the research arm of Tribeck Inc. (conducted on 196 websites of leading companies), newly surveyed AI-search usage in work-related information gathering: about 65% overall already use AI search, and by product/service category, nearly 80% use it in fields such as information systems and energy/environment solutions (PRTIMES #1295890). [F-009] Note: respondent counts and attributes are not stated in this release; "about 65%" and "nearly 80%" are the publisher's approximate wordings.

This site's ANK-2026-07-01-004 records the same shift on the workplace side: generative AI tools inside companies forming a "Big Three Era" split among ChatGPT, Gemini and Copilot, with AI cost moving toward "management-agenda status." That card is the workplace-side cross-section; this card is the consumer-side one — "AI becoming the information entrance" is not a single-industry phenomenon but a structural shift proceeding in sync on both the consumer and workplace sides. The Ahrefs power map cited above adds the supply-side implication: as consumers and companies alike start asking AI for "optimal answers," the visibility battleground for publishers is moving from search rankings to "being cited by AI."

Risk factors

FAQ

Q: What is the "third turning point" in fashion information sources?

Per Onward HD's official conclusion: after "magazines to SNS," fashion information gathering is entering a third wave — "from SNS toward AI, official sources and EC reviews" — and its essence is a change in what consumers seek, from "gazing at aspiration and chasing trends" to "obtaining the optimal answer for oneself."

"Third turning point" and "third wave" are the publisher's framing, reproduced as-is. Supporting data: versus 5 years earlier, SNS +16.2 percentage points, magazines -7.3 points, and AI going from a nonexistent answer option to 17.1% usage; the top purpose is "finding the optimal answer" at 46.4%, with "following admired styles" at only 12.2% (PRTIMES #1296472).

Q: How widespread is SNS fatigue in Japan?

In Onward HD's survey, SNS fatigue in fashion information gathering is 47.2% (often 12.5% + sometimes 34.7%); in Cross Marketing's same-month 3,000-person survey, SNS fatigue among SNS users is 45% (often 11% + sometimes 34%). The questions and populations differ, so the figures are not directly comparable — but they point the same way.

The former is limited to "fatigue, weariness, limits" in fashion information gathering (614 respondents); the latter is defined as having wanted to "keep some distance or take a break" (3,000 people nationwide aged 20-69, June 2026), somewhat more common among people in their 20s-30s (PRTIMES #1296472, PRTIMES #1297146).

Q: Why are people tired of SNS?

The biggest reason is "too much information, hard to find what suits me" at 61.7%, followed by "too many ads and PR, cannot trust it" at 47.2% and "trust in influencers has fallen" at 44.1% (Onward HD survey, fatigued segment of 290 people) — the roots concentrate in "volume" and "trust."

The source notes a structural contradiction: the more SNS expands, the harder it becomes to find information that suits oneself. In Cross Marketing's survey the top scenes are likewise "too much information" 40% and "the same information over and over" 30% (PRTIMES #1296472, PRTIMES #1297146).

Q: Are the SNS-fatigued really moving to AI?

Yes — in the June 2026 survey, among the fatigued segment (290 people), "started using AI more" was the most common change at 37.6%; the fatigued segment's AI-usage growth is 49.3%, 2.4 times the non-fatigued segment's 20.6% (the source's own wording). SNS fatigue is the strongest push behind the shift to AI.

Overall in the June 2026 survey, 32.4% (increased somewhat 22.5% + increased greatly 9.9%) expanded their AI-usage frequency versus 1 year earlier. But the source itself concedes: intent to use AI remains limited, and establishing trustworthiness and usability is the next challenge (PRTIMES #1296472).

Q: How many people now use AI to gather fashion information?

17.1% of Onward HD's survey respondents — a question option that did not exist 5 years earlier. This is a percentage of the survey's pool of 614 people, not a census of all Japanese consumers.

In the same survey, de-SNS information behavior (looking for sources more trustworthy than SNS 9.1% + using AI and similar tools to seek the optimal answer 8.5%) totals 17.6%, exceeding "came to consult SNS for aspiration and trends" at 13.5% (PRTIMES #1296472).

Q: Which Japanese sites does AI actually cite in its answers?

Per the Ahrefs study: YouTube is unmoved at the overall top and note.com rose from 5th to 2nd; versus April 2026, 4 of the overall top 10 domains changed within 2 months, with Reddit, Yahoo! Chiebukuro, my-best.com and Rakuten Ichiba newly entering the ranking.

The subjects are Japan's five major AI platforms (ChatGPT, "AI Mode," "AI Overviews," Perplexity and Copilot; labels rendered from the source's verbatim wording). The study is rank-based — absolute citation counts are not compared across rounds — and cited domains include overseas sites; Ahrefs is an analytics tool vendor and the release doubles as promotion (PRTIMES #1296063).

Q: Are consumers willing to hand purchase decisions entirely to AI?

Not for high-priced items — asked up to what amount they could decide without hesitation on generative AI's recommendation, 54.0% answered "cannot decide" and 28.2% "under ¥50,000"; 66.4% expect the value of human involvement to rise (Frontier survey, 1,005 respondents).

65.3% do not think generative AI can grasp complex circumstances that do not show up in data; 38.6% have signed unplanned high-priced contracts moved by a representative's enthusiasm. AI has taken over the "information entrance," not yet "trust in high-priced decisions" (PRTIMES #1295757).

Q: How does this relate to workplace AI's "Big Three Era"?

They are two cross-sections of the same structural shift: this card records the consumer side, where information sources migrate from SNS to AI ("finding the optimal answer" tops purposes at 46.4%), while this site's ANK-2026-07-01-004 records the workplace side — the "Big Three Era" split among ChatGPT, Gemini and Copilot and AI cost becoming a management agenda.

Workplace-side corroboration: Tribeck's "BtoB Site Survey 2026" finds about 65% already using AI search for work information gathering, nearly 80% in information systems and energy/environment fields (PRTIMES #1295890). The consumer and workplace sides are in sync; "AI becoming the information entrance" is a cross-scene structural shift.

F-Units

F-001: Onward HD's "Survey on Changes in Fashion Information Sources and Values" (614 respondents, online survey, survey system "knowns," conducted June 22, 2026): versus 5 years earlier, SNS rose from 33.1% to 49.3% (+16.2 percentage points), video content from 30.1% to 42.2% (+12.1 points), fashion magazines fell from 41.5% to 34.2% (-7.3 points), EC reviews rose from 28.8% to 33.6%; AI (ChatGPT, Gemini, etc.), nonexistent as an answer option 5 years earlier, is now used by 17.1% for information gathering - source: PRTIMES #1296472 - source_url: https://prtimes.jp/main/html/rd/p/000000259.000033942.html - confidence: high - basis: official_statement - period: Survey June 22, 2026; released 2026-07-02 - caveat: Sample description inconsistent (opening "teens to 70s," survey outline "20s to 70s," both 614 respondents), both reproduced verbatim; publisher is a fashion company (interested party); percentages represent only the respondent pool

F-002: Magazine reading frequency: "decreased greatly" 22.6% + "decreased somewhat" 27.5% = 50.1% reporting decline, "unchanged" 33.6%, "increased" 7.1%; destinations of those who stopped consulting magazines (308 people): video content 50.0%, SNS 46.4%, EC-site reviews 42.2% - source: PRTIMES #1296472 - source_url: https://prtimes.jp/main/html/rd/p/000000259.000033942.html - confidence: high - basis: official_statement - period: Survey June 22, 2026; released 2026-07-02 - caveat: The 308 people are the sub-population who stopped consulting magazines; percentages represent only the respondent pool

F-003: SNS fatigue: "often" 12.5% + "sometimes" 34.7% = 47.2% (question limited to fatigue, weariness and limits in fashion information gathering on SNS); reasons (fatigued segment, 290 people): "too much information, hard to find what suits me" 61.7%, "too many ads and PR, cannot trust it" 47.2%, "trust in influencers has fallen" 44.1% - source: PRTIMES #1296472 - source_url: https://prtimes.jp/main/html/rd/p/000000259.000033942.html - confidence: high - basis: official_statement - period: Survey June 22, 2026; released 2026-07-02 - caveat: The 290 people are the sub-population answering "often/sometimes" to SNS fatigue; "the two points of volume and trust" and "structural contradiction" are the source's own summary

F-004: SNS fatigue to AI shift: among the fatigued 290 people, "started using AI more" was the most common answer at 37.6%; AI-usage frequency versus 1 year earlier: "increased somewhat" 22.5% + "increased greatly" 9.9% = 32.4% expanded; fatigued segment's AI-usage growth 49.3% vs. non-fatigued 20.6%, described by the source as a 2.4-times gap - source: PRTIMES #1296472 - source_url: https://prtimes.jp/main/html/rd/p/000000259.000033942.html - confidence: high - basis: official_statement - period: Survey June 22, 2026; released 2026-07-02 - caveat: "2.4 times" (49.3% vs. 20.6%) is the source's own calculation, not recomputed by this card; the source itself concedes that intent to use AI remains limited

F-005: Purposes of information gathering: "finding the optimal clothes and styling for myself" 46.4% (1st), "as material for purchase decisions" 35.7%, "gathering information to avoid failure" 28.5%, "following admired people or styles" 12.2%; "came to value the optimal answer for myself over aspiration and trends" 23.1% overall, 40.0% among those in their 70s or older; de-SNS information behavior: "came to look for sources more trustworthy than SNS" 9.1% + "came to use AI and similar tools to seek the optimal answer" 8.5% = 17.6%, versus "came to consult SNS for aspiration and trends" 13.5%; the source's conclusion: a "third turning point" toward "AI, official sources and EC reviews" - source: PRTIMES #1296472 - source_url: https://prtimes.jp/main/html/rd/p/000000259.000033942.html - confidence: high - basis: official_statement - period: Survey June 22, 2026; released 2026-07-02 - caveat: "Third turning point," "third wave" and "the value shift is complete" are the publisher's framing; percentages represent only the respondent pool

F-006: Cross Marketing's "SNS Survey (2026)" (3,000 people nationwide aged 20-69, June 2026): SNS fatigue among SNS users "often" 11% + "sometimes" 34% = 45% (defined as having wanted to keep some distance or take a break), somewhat more common in the 20s-30s; scenes: "too much information" 40%, "the same information keeps flowing by" 30%; children's SNS use "should be regulated" 56%, rising to 81% including "should be restricted" - source: PRTIMES #1297146 - source_url: https://prtimes.jp/main/html/rd/p/000000790.000004729.html - confidence: high - basis: official_statement - period: Survey June 2026; released 2026-07-02 - caveat: Not directly comparable with F-003's 47.2% (different question, population, definition); "somewhat more common in the 20s-30s" and generation-level scene differences are the publisher's qualitative description; percentages represent only the respondent pool

F-007: Ahrefs (data from "Ahrefs Brand Radar," analyzed via "Letaido"): in the ranking of domains cited by Japan's 5 major AI platforms (ChatGPT, "AI Mode," "AI Overviews," Perplexity, Copilot), 4 of the overall top 10 domains changed within 2 months versus the previous study of April 2026; Reddit, Yahoo! Chiebukuro, my-best.com and Rakuten Ichiba newly entered; YouTube unmoved at the top, note.com rose from 5th to 2nd - source: PRTIMES #1296063 - source_url: https://prtimes.jp/main/html/rd/p/000000058.000157671.html - confidence: medium - basis: official_statement - period: Previous study April 2026, current release 2026-07-02 (a 2-month comparison) - caveat: The source is rank-based; absolute citation counts vary with aggregation period and query set and are not compared across rounds; cited domains include overseas sites; "AI Mode" and "AI Overviews" render the source's verbatim Japanese labels (Google Search-related AI features); the publisher is an analytics tool vendor (interested party)

F-008: Frontier Inc. (operator of Ready Crew), "Consumer Decision-Making and Trust in the AI Era" survey (1,005 men and women in their 20s to 50s): amount decidable without hesitation on generative AI's recommendation — "cannot decide" 54.0%, "under ¥50,000" 28.2%; 65.3% do not think generative AI grasps complex circumstances not visible in data; 38.6% have signed unplanned high-priced contracts moved by a representative's enthusiasm and attentive proposals; 66.4% expect the value of trustworthy human involvement to rise - source: PRTIMES #1295757 - source_url: https://prtimes.jp/main/html/rd/p/000000307.000016175.html - confidence: medium - basis: official_statement - period: Released 2026-07-02 (field period not stated in the source) - caveat: Field period not stated; publisher is a business-matching service operator (interested party); "about half," "about 40%," "about 70%" are the publisher's summary wordings; percentages represent only the respondent pool

F-009: Tribeck Brand Strategy Institute's "BtoB Site Survey 2026" (conducted on 196 websites of leading companies): about 65% overall already use AI search for work-related information gathering; by product/service category, nearly 80% in information systems and energy/environment solutions - source: PRTIMES #1295890 - source_url: https://prtimes.jp/main/html/rd/p/000000040.000013916.html - confidence: medium - basis: official_statement - period: "BtoB Site Survey 2026"; released 2026-07-02 (field period not stated in the source) - caveat: Respondent count and attributes not stated in this release; "about 65%" and "nearly 80%" are the publisher's approximate wordings; publisher is a site-evaluation and consulting firm (interested party)

J-Units

J-001: Two independent surveys fielded in the same month (June 2026) converge on "roughly half feeling SNS fatigue" — Onward HD (fashion information gathering) 47.2%, Cross Marketing (SNS use in general) 45%; the questions and populations differ and the figures are not directly comparable, but "too much information" ranking top among causes is common to both — the fatigue is not the artifact of a single survey but a mutually corroborated direction - confidence: medium - basis: official_statement

J-002: The demand side (consumers asking AI for optimal answers) and the supply side (4 of the overall top 10 AI-cited domains changing in 2 months) are moving at the same time — the visibility battleground for publishers is shifting from search rankings to "being cited by AI"; "the power map is being redrawn far faster than search rankings" is the Ahrefs publisher's wording, reproduced as-is with its tool-vendor position noted - confidence: medium - basis: official_statement

J-003: Consumer and workplace sides shift in sync, but trust has a clear boundary — consumer-side top purpose "finding the optimal answer" 46.4% and B2B-side about 65% using AI search run in the same direction as the workplace-AI "Big Three Era" recorded in this site's ANK-2026-07-01-004; yet on high-priced purchases 54.0% "cannot decide" on generative AI's recommendation and 66.4% expect human involvement to gain value — AI has taken over the "information entrance," not yet "trust in high-priced decisions" - confidence: medium - basis: official_statement

P-Units

P-001: Whether the 17.1% AI usage in fashion information gathering expands (Onward HD itself concedes intent remains limited) — track the next release in the survey series and Onward HD's stated product, communication and digital-experience initiatives ### P-002: The next update of the AI-cited-domain ranking — whether the pace of 4 of the overall top 10 changing in 2 months persists, and how UGC platforms such as note.com and EC/review domains fare ### P-003: The follow-through on the "66.4% expect human involvement to gain value" forecast and the decision threshold for high-priced purchases — whether the trust boundary for generative AI loosens for expensive items; policy movement on regulating children's SNS use (should be regulated 56%, should be restricted combined 81%)

同事件・三視角 / Three Perspectives on the Same Event / 同一イベント・三つの視点

Internal Citation Chain

Published ANK-Docs cited in this document: - ANK-2026-07-01-004 (Japan's workplace AI enters a "Big Three Era" as cost becomes a management agenda) → cited as the workplace-side counterpart: that card records the structural shift in in-company generative AI tools (a workplace split among ChatGPT, Gemini and Copilot) and AI-cost governance, while this card records the synchronized consumer-side migration of information sources from SNS to AI — together, the two cards are the workplace-side and consumer-side cross-sections of the same "AI becomes the information entrance" structural shift.

Sources

1. [PRTIMES #1296472] 株式会社オンワードホールディングス, "【ファッション情報源もメンパ重視に】SNSに「疲れた」層がAIに移行　「流行・憧れ」でなく「最適解」を求める意識へのシフトが後押し", 2026-07-02. https://prtimes.jp/main/html/rd/p/000000259.000033942.html 2. [PRTIMES #1297146] 株式会社クロス・マーケティング, "SNS疲れを感じるシーンは「情報量の多さ」「同じ情報ばかり」　子どものSNS利用を規制すべきは56％、制限すべきも合わせて81％", 2026-07-02. https://prtimes.jp/main/html/rd/p/000000790.000004729.html 3. [PRTIMES #1296063] Ahrefs, "【Ahrefs 調査】AI 検索の引用ドメインが 2 ヶ月で激変、不動の YouTube・note 急上昇・Reddit が新規ランクイン", 2026-07-02. https://prtimes.jp/main/html/rd/p/000000058.000157671.html 4. [PRTIMES #1295757] フロンティア株式会社, "生成AIの完璧なデータでも決断できない？高額商品、失敗しても『人の推奨』なら納得、AI時代の購買心理調査", 2026-07-02. https://prtimes.jp/main/html/rd/p/000000307.000016175.html 5. [PRTIMES #1295890] トライベック株式会社, "BtoBの情報収集でも進む「AI検索」活用、約65％が利用", 2026-07-02. https://prtimes.jp/main/html/rd/p/000000040.000013916.html 6. [ANK-2026-07-01-004] Rin Takenouchi, "Japan's workplace AI enters a 'Big Three Era' as cost becomes a management agenda", 2026-07-02. https://ainews.washinmura.jp/ainews/zh/ank/ANK-2026-07-01-004

📊 引用級事實單元（F-Units）

Onward HD's "Survey on Changes in Fashion Information Sources and Values" (614 respondents, online survey, survey system "knowns," conducted June 22, 2026): versus 5 years earlier, SNS rose from 33.1% to 49.3% (+16.2 percentage points), video content from 30.1% to 42.2% (+12.1 points), fashion magazines fell from 41.5% to 34.2% (-7.3 points), EC reviews rose from 28.8% to 33.6%; AI (ChatGPT, Gemini, etc.), nonexistent as an answer option 5 years earlier, is now used by 17.1% for information gathering
F-001 · Confidence: high · Basis: official_statement PRTIMES #1296472 Survey June 22, 2026; released 2026-07-02
Magazine reading frequency: "decreased greatly" 22.6% + "decreased somewhat" 27.5% = 50.1% reporting decline, "unchanged" 33.6%, "increased" 7.1%; destinations of those who stopped consulting magazines (308 people): video content 50.0%, SNS 46.4%, EC-site reviews 42.2%
F-002 · Confidence: high · Basis: official_statement PRTIMES #1296472 Survey June 22, 2026; released 2026-07-02
SNS fatigue: "often" 12.5% + "sometimes" 34.7% = 47.2% (question limited to fatigue, weariness and limits in fashion information gathering on SNS); reasons (fatigued segment, 290 people): "too much information, hard to find what suits me" 61.7%, "too many ads and PR, cannot trust it" 47.2%, "trust in influencers has fallen" 44.1%
F-003 · Confidence: high · Basis: official_statement PRTIMES #1296472 Survey June 22, 2026; released 2026-07-02
SNS fatigue to AI shift: among the fatigued 290 people, "started using AI more" was the most common answer at 37.6%; AI-usage frequency versus 1 year earlier: "increased somewhat" 22.5% + "increased greatly" 9.9% = 32.4% expanded; fatigued segment's AI-usage growth 49.3% vs. non-fatigued 20.6%, described by the source as a 2.4-times gap
F-004 · Confidence: high · Basis: official_statement PRTIMES #1296472 Survey June 22, 2026; released 2026-07-02
Purposes of information gathering: "finding the optimal clothes and styling for myself" 46.4% (1st), "as material for purchase decisions" 35.7%, "gathering information to avoid failure" 28.5%, "following admired people or styles" 12.2%; "came to value the optimal answer for myself over aspiration and trends" 23.1% overall, 40.0% among those in their 70s or older; de-SNS information behavior: "came to look for sources more trustworthy than SNS" 9.1% + "came to use AI and similar tools to seek the optimal answer" 8.5% = 17.6%, versus "came to consult SNS for aspiration and trends" 13.5%; the source's conclusion: a "third turning point" toward "AI, official sources and EC reviews"
F-005 · Confidence: high · Basis: official_statement PRTIMES #1296472 Survey June 22, 2026; released 2026-07-02
Cross Marketing's "SNS Survey (2026)" (3,000 people nationwide aged 20-69, June 2026): SNS fatigue among SNS users "often" 11% + "sometimes" 34% = 45% (defined as having wanted to keep some distance or take a break), somewhat more common in the 20s-30s; scenes: "too much information" 40%, "the same information keeps flowing by" 30%; children's SNS use "should be regulated" 56%, rising to 81% including "should be restricted"
F-006 · Confidence: high · Basis: official_statement PRTIMES #1297146 Survey June 2026; released 2026-07-02
Ahrefs (data from "Ahrefs Brand Radar," analyzed via "Letaido"): in the ranking of domains cited by Japan's 5 major AI platforms (ChatGPT, "AI Mode," "AI Overviews," Perplexity, Copilot), 4 of the overall top 10 domains changed within 2 months versus the previous study of April 2026; Reddit, Yahoo! Chiebukuro, my-best.com and Rakuten Ichiba newly entered; YouTube unmoved at the top, note.com rose from 5th to 2nd
F-007 · Confidence: medium · Basis: official_statement PRTIMES #1296063 Previous study April 2026, current release 2026-07-02 (a 2-month comparison)
Frontier Inc. (operator of Ready Crew), "Consumer Decision-Making and Trust in the AI Era" survey (1,005 men and women in their 20s to 50s): amount decidable without hesitation on generative AI's recommendation — "cannot decide" 54.0%, "under ¥50,000" 28.2%; 65.3% do not think generative AI grasps complex circumstances not visible in data; 38.6% have signed unplanned high-priced contracts moved by a representative's enthusiasm and attentive proposals; 66.4% expect the value of trustworthy human involvement to rise
F-008 · Confidence: medium · Basis: official_statement PRTIMES #1295757 Released 2026-07-02 (field period not stated in the source)
Tribeck Brand Strategy Institute's "BtoB Site Survey 2026" (conducted on 196 websites of leading companies): about 65% overall already use AI search for work-related information gathering; by product/service category, nearly 80% in information systems and energy/environment solutions
F-009 · Confidence: medium · Basis: official_statement PRTIMES #1295890 "BtoB Site Survey 2026"; released 2026-07-02 (field period not stated in the source)

❓ FAQ

What is the "third turning point" in fashion information sources?

Per Onward HD's official conclusion: after "magazines to SNS," fashion information gathering is entering a third wave — "from SNS toward AI, official sources and EC reviews" — and its essence is a change in what consumers seek, from "gazing at aspiration and chasing trends" to "obtaining the optimal answer for oneself." "Third turning point" and "third wave" are the publisher's framing, reproduced as-is. Supporting data: versus 5 years earlier, SNS +16.2 percentage points, magazines -7.3 points, and AI going from a nonexistent answer option to 17.1% usage; the top purpose is "finding the optimal answer" at 46.4%, with "following admired styles" at only 12.2% (PRTIMES #1296472).

How widespread is SNS fatigue in Japan?

In Onward HD's survey, SNS fatigue in fashion information gathering is 47.2% (often 12.5% + sometimes 34.7%); in Cross Marketing's same-month 3,000-person survey, SNS fatigue among SNS users is 45% (often 11% + sometimes 34%). The questions and populations differ, so the figures are not directly comparable — but they point the same way. The former is limited to "fatigue, weariness, limits" in fashion information gathering (614 respondents); the latter is defined as having wanted to "keep some distance or take a break" (3,000 people nationwide aged 20-69, June 2026), somewhat more common among people in their 20s-30s (PRTIMES #1296472, PRTIMES #1297146).

Why are people tired of SNS?

The biggest reason is "too much information, hard to find what suits me" at 61.7%, followed by "too many ads and PR, cannot trust it" at 47.2% and "trust in influencers has fallen" at 44.1% (Onward HD survey, fatigued segment of 290 people) — the roots concentrate in "volume" and "trust." The source notes a structural contradiction: the more SNS expands, the harder it becomes to find information that suits oneself. In Cross Marketing's survey the top scenes are likewise "too much information" 40% and "the same information over and over" 30% (PRTIMES #1296472, PRTIMES #1297146).

Are the SNS-fatigued really moving to AI?

Yes — in the June 2026 survey, among the fatigued segment (290 people), "started using AI more" was the most common change at 37.6%; the fatigued segment's AI-usage growth is 49.3%, 2.4 times the non-fatigued segment's 20.6% (the source's own wording). SNS fatigue is the strongest push behind the shift to AI. Overall in the June 2026 survey, 32.4% (increased somewhat 22.5% + increased greatly 9.9%) expanded their AI-usage frequency versus 1 year earlier. But the source itself concedes: intent to use AI remains limited, and establishing trustworthiness and usability is the next challenge (PRTIMES #1296472).

How many people now use AI to gather fashion information?

17.1% of Onward HD's survey respondents — a question option that did not exist 5 years earlier. This is a percentage of the survey's pool of 614 people, not a census of all Japanese consumers. In the same survey, de-SNS information behavior (looking for sources more trustworthy than SNS 9.1% + using AI and similar tools to seek the optimal answer 8.5%) totals 17.6%, exceeding "came to consult SNS for aspiration and trends" at 13.5% (PRTIMES #1296472).

Which Japanese sites does AI actually cite in its answers?

Per the Ahrefs study: YouTube is unmoved at the overall top and note.com rose from 5th to 2nd; versus April 2026, 4 of the overall top 10 domains changed within 2 months, with Reddit, Yahoo! Chiebukuro, my-best.com and Rakuten Ichiba newly entering the ranking. The subjects are Japan's five major AI platforms (ChatGPT, "AI Mode," "AI Overviews," Perplexity and Copilot; labels rendered from the source's verbatim wording). The study is rank-based — absolute citation counts are not compared across rounds — and cited domains include overseas sites; Ahrefs is an analytics tool vendor and the release doubles as promotion (PRTIMES #1296063).

Are consumers willing to hand purchase decisions entirely to AI?

Not for high-priced items — asked up to what amount they could decide without hesitation on generative AI's recommendation, 54.0% answered "cannot decide" and 28.2% "under ¥50,000"; 66.4% expect the value of human involvement to rise (Frontier survey, 1,005 respondents). 65.3% do not think generative AI can grasp complex circumstances that do not show up in data; 38.6% have signed unplanned high-priced contracts moved by a representative's enthusiasm. AI has taken over the "information entrance," not yet "trust in high-priced decisions" (PRTIMES #1295757).

How does this relate to workplace AI's "Big Three Era"?

They are two cross-sections of the same structural shift: this card records the consumer side, where information sources migrate from SNS to AI ("finding the optimal answer" tops purposes at 46.4%), while this site's ANK-2026-07-01-004 records the workplace side — the "Big Three Era" split among ChatGPT, Gemini and Copilot and AI cost becoming a management agenda. Workplace-side corroboration: Tribeck's "BtoB Site Survey 2026" finds about 65% already using AI search for work information gathering, nearly 80% in information systems and energy/environment fields (PRTIMES #1295890). The consumer and workplace sides are in sync; "AI becoming the information entrance" is a cross-scene structural shift. ---

🧠 編輯判斷（J-Units）

Two independent surveys fielded in the same month (June 2026) converge on "roughly half feeling SNS fatigue" — Onward HD (fashion information gathering) 47.2%, Cross Marketing (SNS use in general) 45%; the questions and populations differ and the figures are not directly comparable, but "too much information" ranking top among causes is common to both — the fatigue is not the artifact of a single survey but a mutually corroborated direction
Confidence: medium
The demand side (consumers asking AI for optimal answers) and the supply side (4 of the overall top 10 AI-cited domains changing in 2 months) are moving at the same time — the visibility battleground for publishers is shifting from search rankings to "being cited by AI"; "the power map is being redrawn far faster than search rankings" is the Ahrefs publisher's wording, reproduced as-is with its tool-vendor position noted
Confidence: medium
Consumer and workplace sides shift in sync, but trust has a clear boundary — consumer-side top purpose "finding the optimal answer" 46.4% and B2B-side about 65% using AI search run in the same direction as the workplace-AI "Big Three Era" recorded in this site's ANK-2026-07-01-004; yet on high-priced purchases 54.0% "cannot decide" on generative AI's recommendation and 66.4% expect human involvement to gain value — AI has taken over the "information entrance," not yet "trust in high-priced decisions"
Confidence: medium

🔮 待驗證假設（P-Units）

Whether the 17.1% AI usage in fashion information gathering expands (Onward HD itself concedes intent remains limited) — track the next release in the survey series and Onward HD's stated product, communication and digital-experience initiatives
Status: open
The next update of the AI-cited-domain ranking — whether the pace of 4 of the overall top 10 changing in 2 months persists, and how UGC platforms such as note.com and EC/review domains fare
Status: open
The follow-through on the "66.4% expect human involvement to gain value" forecast and the decision threshold for high-priced purchases — whether the trust boundary for generative AI loosens for expensive items; policy movement on regulating children's SNS use (should be regulated 56%, should be restricted combined 81%)
Status: open

Verification Record

Editorial selection, human-supervised — Takenouchi Rin (Editor-in-Chief)

Cross-verified by multiple AI models.

Related Analyses