AI TV Purchase-Consultation Retailer Recommendation Survey: Which Electronics Retailer Do the Five Major Generative AIs Recommend for Buying a TV?

Summary

To see which purchase sources and electronics retailers major generative AIs recommend for buying a TV, Stellagent Inc. asked five AI services — ChatGPT, Gemini, Claude, Perplexity, and Microsoft Copilot — across six purchase scenarios, five times each, for 150 questions in total. The survey had two stages: first we checked recommendation trends across purchase-source categories including electronics retailers, Amazon, Rakuten Ichiba, Kakaku.com, and manufacturer official; then we limited the question to electronics retailers and, under a nationwide, general-context condition that uses neither the user's location nor prior context, checked which retailer names the AI ranked highly.

Metric	Result
Target AIs	5 services (ChatGPT, Gemini, Claude, Perplexity, Copilot)
Number of runs	5 AIs × 6 scenarios × 5 runs · 150 answers total
Recommendation rows extracted	725
Primary tally (within-retailer comparison)	75 answers (S4R–S6R)
Top on ranking points in the primary tally	Yamada Denki (220 points)
Most first-place picks in the primary tally	Yodobashi Camera (26 times)
Mentions of warranty	72 of 75 answers (96.0%)
Mentions of energy efficiency / electricity cost	0 of 75 answers (0.0%)

In the within-retailer comparison (S4R–S6R, 75 answers), Yamada Denki led with 220 ranking points. Yodobashi Camera and K's Denki tied at 204 points, and Yodobashi Camera had the most first-place picks at 26. The full text of the six prompts used in this survey is provided in the appendix at the end of this article.

The chart above may be reproduced as-is in media coverage, articles, and other materials, provided that you credit "Stellagent Inc." as the source and include a link to this page.

Meanwhile, warranty, in-store consultation, delivery/installation, and recycling collection were mentioned in over 90% of answers, while for TV-specific confirmation items, mentions of HDMI/gaming use came to 4 answers (5.3%) and energy efficiency / electricity cost to 0 answers (0.0%).

Background

Generative AI is becoming an entry point for gathering information and comparison consultation when buying home appliances. A TV involves not only model number and price but also many post-purchase satisfaction factors: screen size, viewing distance, installation location, wall-mount vs. stand, antenna and internet connection, video streaming services, HDMI ports, gaming use, recycling collection of the old TV, and warranty.

For electronics retailers, what matters is not only which brands AI recommends in its answers, but which purchase-support information AI understands and which confirmation items it tends to drop. Whether information — covering not just price comparison and EC marketplaces but also in-store consultation, delivery/installation, wiring, collection, and warranty — reaches AI will affect the future purchase journey.

In this survey, we entered TV purchase consultations into major generative AIs under identical conditions and observed recommendation trends for purchase-source categories and electronics-retailer names, along with how often TV-purchase confirmation items were mentioned.

Scenarios Used

We designed six scenarios in two stages: the purchase-source category comparison (S1–S3) and the within-retailer comparison (S4R–S6R). S4R–S6R were entered as nationwide generalizations that use neither the user's location nor nearby-store information, and were treated as the authoritative data for the published ranking. The full text of the prompts actually entered for each scenario is provided in the appendix at the end of this article.

ID	Scenario	Condition
S1	General TV purchase consultation	No purchase-source category specified; observe which sources the AI naturally raises as a baseline
S2	Purchase consultation with installation/connection concerns	State living-room use and concerns about screen size, viewing distance, installation location, wall-mount vs. stand, antenna/internet connection, and collection of the old TV; observe mentions of installation/wiring/collection-included purchase and pre-purchase confirmation items
S3	Price-focused, comparing online purchases	Explicitly compare Amazon, Rakuten, Kakaku.com, electronics retailers, and manufacturer official; see whether it still flags installation/collection/warranty risk even when price comes first
S4R	General comparison within electronics retailers (nationwide)	Without using location, address, IP address, nearby stores, account information, search history, or past conversation, observe top retailer-name recommendations as a nationwide generalization
S5R	Installation/connection-focused comparison within retailers (nationwide)	Without location etc., observe retailer recommendations that account for screen size, installation, wall-mount/stand, antenna/internet connection, recycling collection, and warranty
S6R	Balancing price and peace of mind within retailers (nationwide)	Without location etc., observe retailer recommendations that balance low price with the peace of mind of installation/connection/collection/warranty

Yamada Denki Leads on Ranking Points in the Within-Retailer Comparison

As the authoritative data for the published ranking, we tallied the 75 answers from S4R–S6R. As a result, Yamada Denki led with 220 ranking points.

Rank	Normalized name	Ranking points	Recommendations	First-place picks	AIs spanned	Scenarios spanned
1	Yamada Denki	220	69	9	5	3
2	Yodobashi Camera	204	60	26	5	3
3	K's Denki	204	67	15	5	3
4	EDION	167	46	17	5	3
5	Bic Camera	155	58	2	5	3
6	Nojima	57	23	4	5	3
7	Online retailers	40	12	2	2	3
8	Kojima	20	10	0	3	3
9	Rakuten Ichiba	20	7	0	3	3
10	Joshin	5	4	0	2	3

Ranking points are tallied as 5 points for first place, 4 for second, 3 for third, 2 for fourth, and 1 for fifth. In a tie, we compare in the order of first-place picks, recommendation count, number of AI services spanned, and number of scenarios spanned. Yodobashi Camera and K's Denki are both at 204 points, but because first-place picks are 26 for Yodobashi Camera and 15 for K's Denki, this table places Yodobashi Camera higher. On recommendation count, K's Denki exceeds Yodobashi Camera, 67 to 60.

Yamada Denki had only 9 first-place picks, but its recommendation count of 69 was the highest. Within the scope of this survey, appearing steadily as a candidate across all AIs and all scenarios led to its first-place finish on ranking points. Yodobashi Camera, meanwhile, had the most first-place picks at 26. In AI answers, it tended to rise in contexts such as product selection, hands-on comparison, point rewards, delivery/installation, and in-store consultation.

By Scenario, the Leaders Split Across General, Installation-Focused, and Price-and-Peace-of-Mind

In the three within-retailer scenarios, the leader differed by scenario.

Scenario	1st	2nd	3rd	Trend
S4R General comparison	K's Denki (86 pts)	Yodobashi Camera (72 pts)	Yamada Denki (62 pts)	K's Denki rose in the context of long-term warranty and cash discounts.
S5R Installation/connection-focused	Yodobashi Camera (80 pts)	Yamada Denki (69 pts)	EDION (59 pts)	Yodobashi Camera rose in the context of installation, collection, in-store consultation, and hands-on comparison.
S6R Balancing price and peace of mind	Yamada Denki (89 pts)	K's Denki (60 pts)	Bic Camera (58 pts)	Yamada Denki rose in the scenario asking to balance low price with warranty and installation.

S4R–S6R were all entered as nationwide generalizations that do not use location or nearby-store information. The results therefore reflect not specific-area store density but the brand names and purchase-support elements AI tends to recall in a general TV purchase consultation.

The Brands That Rose to the Top Differed Sharply by AI

By AI, ChatGPT and Copilot strongly favored Yodobashi Camera, Gemini favored Yamada Denki, Claude favored K's Denki and EDION, and Perplexity favored EDION.

AI service	1st	2nd	3rd	Note
ChatGPT	Yodobashi Camera (61 pts)	K's Denki (47 pts)	Bic Camera (46 pts)	Yodobashi Camera had 8 first-place picks
Gemini	Yamada Denki (61 pts)	Bic Camera (49 pts)	K's Denki (46 pts)	Yamada Denki had 7 first-place picks
Claude	K's Denki (39 pts)	EDION (39 pts)	Yodobashi Camera (37 pts)	K's Denki and EDION tied. First-place picks: K's Denki 7, EDION 5
Perplexity	EDION (72 pts)	Yamada Denki (59 pts)	K's Denki (43 pts)	EDION had 12 first-place picks
Copilot	Yodobashi Camera (70 pts)	EDION (46 pts)	Yamada Denki (42 pts)	Yodobashi Camera had 14 first-place picks

Even with the same prompt, the brands that rose to the top varied greatly by AI service. Within the scope of this survey, this shows that a single search ranking or ad exposure alone cannot fully explain how a brand appears in AI answers.

Warranty, In-Store Consultation, Delivery/Installation, and Collection Exceed 90%; HDMI/Gaming Use and Energy Efficiency Are Easily Missed

For the 75 answers in the within-retailer comparison, we tallied mentions of TV-purchase confirmation items on a per-answer basis.

Confirmation item	Answers with a mention	Mention rate
Warranty	72	96.0%
In-store / in-person consultation	71	94.7%
Delivery / installation	69	92.0%
Recycling collection	69	92.0%
Pickup / collection of the old TV	68	90.7%
Antenna / broadcasting	53	70.7%
Internet connection / video streaming	53	70.7%
Wall-mount / stand	51	68.0%
On-site inspection / estimate	44	58.7%
Delivery/installation-included purchase	34	45.3%
Screen size / viewing distance	30	40.0%
Official EC link	23	30.7%
Audio	10	13.3%
HDMI / gaming use	4	5.3%
Energy efficiency / electricity cost	0	0.0%

The chart above may be reproduced as-is in media coverage, articles, and other materials, provided that you credit "Stellagent Inc." as the source and include a link to this page.

AI answers touched frequently on purchase-support items such as warranty, in-store consultation, delivery/installation, and recycling collection. By contrast, HDMI/gaming use — tied to TV-specific use cases — came to only 4 answers (5.3%) and audio to 10 answers (13.3%), while energy efficiency / electricity cost was 0 answers (0.0%). This result suggests that while AI readily recalls in-store consultation, installation/collection, and warranty as "reasons to buy a TV at an electronics retailer," it may not sufficiently cover confirmation items tailored to the user's environment and viewing style.

In the Purchase-Source Category Comparison, EC and Price-Comparison Sites Rose Under Price-Focused Conditions

In the S1–S3 purchase-source category comparison, electronics retailers led with 200 recommendation rows. EC marketplaces / online retailers had 121 recommendation rows, and price-comparison sites had 24.

Purchase-source category	Recommendation rows extracted
Electronics retailers	200
EC marketplace / online retailer	121
Price-comparison site	24
Manufacturer direct	13
Specialty / installer / other	6
Used / reuse	2

By scenario, in S1 (general TV purchase consultation) and S2 (installation/connection concerns), electronics retailers such as Yodobashi Camera, Bic Camera, and Yamada Denki rose to the top. In the price-focused S3 scenario, Kakaku.com scored 86 points, Amazon 76, and Rakuten Ichiba 64, with EC marketplaces and price-comparison sites rising to the top.

Scenario	1st	2nd	3rd
S1 General TV purchase consultation	Yodobashi Camera (63 pts)	Bic Camera (51 pts)	Yamada Denki (44 pts)
S2 Purchase consultation with installation/connection concerns	Yodobashi Camera (87 pts)	Bic Camera (64 pts)	Yamada Denki (58 pts)
S3 Price-focused, comparing online purchases	Kakaku.com (86 pts)	Amazon (76 pts)	Rakuten Ichiba (64 pts)

When price was strongly specified, AI tended to surface EC marketplaces and price-comparison sites. When concerns about installation, connection, collection, and warranty were made explicit, electronics retailers tended to rise instead.

Non-Specific Answers and Notes on Extraction

Across all 150 answers, we were able to extract some specific purchase-source name or purchase-source category. There were 0 answers from which no specific name could be extracted.

That said, within an answer, category terms such as "electronics retailer," "major EC," and "online retailer" may appear alongside specific names such as Yodobashi Camera, Yamada Denki, and Amazon. In this survey, we extracted whatever appeared in the body text as a purchase source on a per-recommendation basis and tallied it after normalizing notation variants; the number of extracted rows therefore exceeds the number of answers. Some Claude and Perplexity answers contained what appeared to be externally sourced citation names or fragments; these were saved as AI answer logs but were not adopted as official fact in the results in the body.

Implications for Businesses

Purchase-Support Information Read by AI Is Not Just About Price

In this survey, the reasons electronics retailers were recommended frequently included in-store consultation, delivery/installation, recycling collection, and warranty. In TV sales, it is important to organize purchase-support information — not just product price and points, but installation, wiring, collection, and warranty — on official sites and FAQs in a form AI can easily understand.

TV-Specific Use-Case Information Is Easily Missed in AI Answers

Mentions of HDMI/gaming use, audio, and energy efficiency / electricity cost were low, and energy efficiency / electricity cost in particular was 0 answers in this primary tally. For users choosing a TV, game-console connection, video streaming, audio, power consumption, viewing distance, and screen size are important comparison axes. Structuring this information across product pages, buying-guide articles, FAQs, and in-store consultation flows creates room for it to be picked up in AI answers.

Because Appearances Differ by AI, Checking a Single AI Is Not Enough

The brands that rose to the top differed across ChatGPT, Gemini, Claude, Perplexity, and Copilot. When checking exposure in AI answers, one should not judge from a single AI's answer but observe trends across multiple AIs, multiple scenarios, and multiple runs.

Nationwide Generalizations and Area-Specific Recommendations Must Be Designed Separately

In S4R–S6R, we asked as a nationwide generalization that does not use location or nearby-store information. In actual purchasing, nearby stores, delivery-coverage areas, installation-coverage areas, stock, and campaigns shape the outcome. On official sites, it is important to organize separately the strengths conveyed as a nationwide generalization and the information that should be conveyed by area and by store.

Survey Method (Details for Reproduction)

This is an AI-answer experiment. So readers can reproduce and verify it under the same conditions, we disclose the operator, subjects, selection criteria, target AIs and conditions, context-influence control, recorded items, extraction/normalization/aggregation methods, deviations/reruns/exclusions, external sources, and the handling of ethics and privacy.

The objective was to check which purchase sources and electronics retailers major generative AIs recommend for a TV purchase consultation, and whether they guide users all the way to TV-purchase confirmation items. The conducting body is Stellagent Inc., and the operator is Codex. The subjects are recommendation answers by major generative AIs about where to buy a TV, and the survey date is August 3, 2026. The selection criteria for the target AIs were major general-purpose generative AI services that general users in Japan can use via a web UI and that are commonly used for product purchase consultations. We included search-engine-type and answer-engine-type services to compare differences in purchase-source recommendation exposure. The number of answers collected is 5 AIs × 6 scenarios × 5 runs each, 150 answers in total.

Target AIs and Conditions

For each answer, we saved, per answer, the collection timestamp, AI service, model name or display name, contract plan, web-search setting, scenario ID, run number, input prompt, retrieved URL, and full answer text. The conditions for each target AI are as follows.

AI service	Model name / display name	Contract plan	Web-search setting
ChatGPT	ChatGPT Web (model display unconfirmed)	ChatGPT Plus (user's contracted environment)	No explicit search operation (answers may include web-sourced citation displays)
Gemini	Gemini Flash	Logged in, Pro display	Automatic web search used
Claude	Claude Opus 5, high	Pro plan	Web-search setting unspecified
Perplexity	Perplexity	Free plan	Search mode
Copilot	Copilot Smart	Logged in	Automatic web search used

Context-Influence Control

Each answer was collected in a new chat, Temporary mode, a temporary chat, an incognito chat, or a chat that can suppress contextual influence. Within the prompt, we also specified not to reference past conversation or memory. In S4R–S6R, we specified not to use current location, address, IP address, nearby stores, nearest stores, local store conditions, account information, search history, past conversation, or memory. This let us observe how retailer names appear as a nationwide generalization for all of Japan, rather than specific-area store conditions.

Extraction, Normalization, and Aggregation

We extracted the purchase sources that appeared in each answer, using their order of appearance in the answer as the recommendation rank. Even when the AI did not state an explicit ranking, we treated the order of appearance in the body text as the recommendation rank. When multiple stores were listed together, we still recorded them in order of appearance in the body text at extraction time.

We saved the original notation as-is, then converted it to a normalized name for aggregation. For example, "Yodobashi" and "Yodobashi.com" were treated as Yodobashi Camera, "Yamada" and "Yamada Denki (Yamada Electric)" as Yamada Denki, and "K's" as K's Denki. Amazon, Rakuten Ichiba, Yahoo! Shopping, online retailers, Kakaku.com, manufacturer official, home centers, specialty stores, and used/reuse were also classified as purchase-source categories.

Ranking points were tallied as 5 points for first place, 4 for second, 3 for third, 2 for fourth, and 1 for fifth. In the within-retailer ranking, we list items in descending order of ranking points, and in a tie compare in the order of first-place picks, recommendation count, number of AI services spanned, and number of scenarios spanned. Mentions of confirmation items were judged on a per-answer basis; because per-recommendation extraction rows duplicate the same answer's confirmation items across multiple rows, mention rates were re-tallied against the 75 answers of S4R–S6R as the denominator. Extraction and normalization were script-assisted, and for the top brands and main confirmation items we re-tallied ranking points, recommendation counts, and mention rates from the primary data and cross-checked them.

Deviations, Reruns, and Exclusions

Perplexity reached the free-search limit on the first collection, so we logged back in with a separate account to collect the shortfall. Claude extracted an in-progress generation UI on the first collection, so we corrected the collection method and re-collected the complete answer text. Short logs that captured only limit screens or in-progress UI were deleted and are not included in the final tally. In the end, there are no missing answers and no short answers. Across all 150 answers, a specific purchase-source name or purchase-source category could be extracted, and there were 0 answers from which no specific name could be extracted.

External Sources and Confirmation Method

External sources were confirmed on August 3, 2026. External sources were used not to judge the correctness of AI answers, but to confirm, when interpreting results, that official flows for purchase-support items exist. This survey does not judge the relative merits of each company's services based on the official pages below. We treated the observed AI answers and the existence of officially confirmable purchase-support flows separately. The main official pages we confirmed are as follows.

Confirmation target	Brand / site	URL	Purpose
TV / recorder installation	Yodobashi Camera	Source	Confirm TV-specific installation flow
Recycling collection	Yodobashi Camera	Source	Confirm home-appliance recycling scope
From TV purchase to installation	Bic Camera	Source	Confirm TV purchase, installation, and recycling notes
Home-appliance recycling	Bic Camera	Source	Confirm recycling fees and target items
Delivery/installation & work	Yamada Denki	Source	Confirm delivery/installation and wall-mount TV installation
Home-appliance recycling	Yamada Denki	Source	Confirm recycling-collection application flow
Large delivery items	K's Denki	Source	Confirm delivery/installation of designated large items including TVs
Recycling collection	K's Denki	Source	Confirm recycling-collection flow
Recycling FAQ	EDION	Source	Confirm unwanted-appliance pickup guidance
Large-item delivery / recycling	Nojima Online	Source	Confirm delivery, installation, and recycling of large items
Recycling FAQ	Joshin	Source	Confirm simultaneous-pickup flow for TVs and similar items

Ethics and Privacy

This is an AI-answer experiment observing generative AIs' public answers, not a survey of human subjects, so no personally identifying information was collected. In S4R–S6R, we explicitly specified in the prompt not to use personal or local context such as current location, address, IP address, account information, or search history, and asked as a nationwide generalization. External site names and citation fragments contained in answers were saved as AI answer logs but were not transcribed into the body as official fact. In the published body, we tallied which purchase sources the AI recommended and which confirmation items it mentioned.

Notes on the Survey

• This is an exploratory AI-answer experiment of 5 AIs × 6 scenarios × 5 runs each, and is not intended for statistical generalization.
• AI answers may vary with date and time, model, contract plan, search settings, conversation history, account state, regional information, and web-search results.
• This survey observed the purchase sources and confirmation items that appeared in AI answers; it does not evaluate the relative merits of each electronics retailer's price, service quality, warranty terms, delivery quality, stock, or in-store support.
• Ranking points are tallied based on order of appearance in the answer or on an explicit ranking. Even when the AI treats multiple candidates as equivalent, we rank them by order of appearance in the body text.
• External sources are official pages confirmed as of August 3, 2026. Each company's delivery/installation, recycling, warranty, and in-store consultation flows are subject to change.
• Product and service names in this report are trademarks or registered trademarks of their respective companies.

Revision History

• August 4, 2026: First published

Prompts Used

The 150 responses used in the published tally were generated with the following six prompts. S1–S3 were used in the purchase-source category comparison, and S4R–S6R in the within-retailer comparison (nationwide generalization). The AI services received the Japanese originals shown below; English translations are provided for reference.

S1

Original Japanese prompt used

テレビを買い替えたいです。どこで買うのがよいか、購入先を3〜5個、おすすめ順に挙げて、理由と注意点を教えてください。過去の会話や記憶は参照せず、この条件だけで回答してください。

English translation

I want to replace my TV. Where would be a good place to buy one? Please list three to five purchase options in recommended order, and explain the reasons and points to note. Do not refer to past conversations or memory; answer based only on these conditions.

S2

Original Japanese prompt used

リビングで使うテレビを買いたいのですが、画面サイズ、視聴距離、設置場所、壁掛けにするかスタンドにするか、アンテナやネット接続、古いテレビのリサイクル回収などが少し不安です。どこで買うのがよいか、購入先を3〜5個、おすすめ順に挙げて、理由と注意点を教えてください。過去の会話や記憶は参照せず、この条件だけで回答してください。

English translation

I want to buy a TV for the living room, but I am a little concerned about screen size, viewing distance, installation location, whether to wall-mount or use a stand, antenna and internet connection, and recycling collection of the old TV. Where would be a good place to buy one? Please list three to five purchase options in recommended order, and explain the reasons and points to note. Do not refer to past conversations or memory; answer based only on these conditions.

S3

Original Japanese prompt used

テレビをできるだけ安く買いたいです。Amazon、楽天、価格.com、家電量販店、メーカー公式など、どこで買うのがよいか迷っています。購入先を3〜5個、おすすめ順に挙げて、理由と注意点を教えてください。過去の会話や記憶は参照せず、この条件だけで回答してください。

English translation

I want to buy a TV as cheaply as possible. I am unsure where to buy it — Amazon, Rakuten, Kakaku.com, an electronics retailer, or the manufacturer's official store, for example. Please list three to five purchase options in recommended order, and explain the reasons and points to note. Do not refer to past conversations or memory; answer based only on these conditions.

S4R

Original Japanese prompt used

テレビを家電量販店で買うなら、どの家電量販店がよいですか。おすすめの家電量販店を3〜5個、おすすめ順に挙げて、理由と注意点を教えてください。現在地、住所、IPアドレス、近隣店舗、最寄り店舗、地域の店舗事情、アカウント情報、検索履歴、過去の会話、記憶は一切使わないでください。東京・江東区など特定地域を前提にせず、日本全国の一般論として回答してください。

English translation

If I were to buy a TV from an electronics retailer, which electronics retailer would be best? Please list three to five recommended electronics retailers in order, and explain the reasons and points to note. Do not use my current location, address, IP address, nearby stores, nearest stores, local store conditions, account information, search history, past conversations, or memory under any circumstances. Do not assume a specific area such as Tokyo or Koto City; answer in general terms for Japan nationwide.

S5R

Original Japanese prompt used

テレビを家電量販店で買いたいです。画面サイズ、視聴距離、設置場所、壁掛け/スタンド、アンテナやネット接続、リサイクル回収、保証まで含めて安心できる家電量販店はどこですか。おすすめ順に3〜5個挙げて、理由と注意点を教えてください。現在地、住所、IPアドレス、近隣店舗、最寄り店舗、地域の店舗事情、アカウント情報、検索履歴、過去の会話、記憶は一切使わないでください。東京・江東区など特定地域を前提にせず、日本全国の一般論として回答してください。

English translation

I want to buy a TV from an electronics retailer. Which electronics retailer can I trust when screen size, viewing distance, installation location, wall-mount/stand, antenna and internet connection, recycling collection, and warranty are all taken into account? Please list three to five in recommended order, and explain the reasons and points to note. Do not use my current location, address, IP address, nearby stores, nearest stores, local store conditions, account information, search history, past conversations, or memory under any circumstances. Do not assume a specific area such as Tokyo or Koto City; answer in general terms for Japan nationwide.

S6R

Original Japanese prompt used

テレビを家電量販店でできるだけ安く、でも設置・アンテナ/ネット接続・リサイクル回収・保証で失敗しないように買いたいです。どの家電量販店がよいですか。おすすめ順に3〜5個挙げて、理由と注意点を教えてください。現在地、住所、IPアドレス、近隣店舗、最寄り店舗、地域の店舗事情、アカウント情報、検索履歴、過去の会話、記憶は一切使わないでください。東京・江東区など特定地域を前提にせず、日本全国の一般論として回答してください。

English translation

I want to buy a TV from an electronics retailer as cheaply as possible, but without failing on installation, antenna/internet connection, recycling collection, or warranty. Which electronics retailer would be best? Please list three to five in recommended order, and explain the reasons and points to note. Do not use my current location, address, IP address, nearby stores, nearest stores, local store conditions, account information, search history, past conversations, or memory under any circumstances. Do not assume a specific area such as Tokyo or Koto City; answer in general terms for Japan nationwide.