[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"article:recruit-data-platform":3},{"meta":4,"markdown":66,"quiz":67},{"type":5,"articleId":6,"slug":7,"title":8,"titleEn":9,"category":10,"summary":11,"publishedAt":12,"image":13,"vocabulary":14,"quizId":65},"article","tech-recruit-data-platform","recruit-data-platform","Recruit データ基盤刷新事例 — 数十サービスの統合","Recruit Data Platform Renewal — Consolidating Dozens of Services","tech","Recruit Group operates dozens of services — SUUMO, Hot Pepper, Townwork, and more. This article describes their data platform consolidation: BigQuery as central warehouse, dbt for transformations, data mesh principles for organizational scaling, and privacy compliance under the Personal Information Protection Act.\n","2026-04-27T00:00:00Z","https:\u002F\u002Fimages.yamiyomi.com\u002Ftech-recruit-data-platform.png",[15,20,25,29,33,37,41,45,49,53,57,61],{"word":16,"reading":17,"meaning":18,"level":19},"刷新","さっしん","renewal","N1",{"word":21,"reading":22,"meaning":23,"level":24},"統合","とうごう","consolidation","N2",{"word":26,"reading":27,"meaning":28,"level":24},"倉庫","そうこ","warehouse",{"word":30,"reading":31,"meaning":32,"level":24},"変換","へんかん","transformation",{"word":34,"reading":35,"meaning":36,"level":19},"系譜","けいふ","lineage",{"word":38,"reading":39,"meaning":40,"level":19},"分権","ぶんけん","decentralization",{"word":42,"reading":43,"meaning":44,"level":24},"個人情報","こじんじょうほう","personal information",{"word":46,"reading":47,"meaning":48,"level":24},"保護","ほご","protection",{"word":50,"reading":51,"meaning":52,"level":19},"遵守","じゅんしゅ","compliance",{"word":54,"reading":55,"meaning":56,"level":24},"規模","きぼ","scale",{"word":58,"reading":59,"meaning":60,"level":24},"重複","ちょうふく","duplication",{"word":62,"reading":63,"meaning":64,"level":24},"領域","りょういき","domain","tech-data-cloud-quiz","\n::para\nリクルートグループは、SUUMO、ホットペッパー、[タウンワーク]{たうんわーく:Townwork}、ゼクシィ、リクナビなど[数十]{すうじゅう:dozens:N3}のサービスを[展開]{てんかい:operate:N1}しています。それぞれが[長年]{ながねん:many years:N5}にわたって[独自]{どくじ:independent:N1}のデータ[基盤]{きばん:platform:N1}を[構築]{こうちく:built:N2}してきた[結果]{けっか:result:N1}、データの[重複]{ちょうふく:duplication:N2}や[整合性]{せいごうせい:consistency:N1}の[問題]{もんだい:problems:N4}が[顕在化]{けんざいか:became apparent:N1}していました。[本]{ほん:this:N5}[記事]{きじ:article:N3}では、リクルートが[数年]{すうねん:several years:N3}[掛けて]{かけて:taking:N3}[実施]{じっし:implemented:N1}した[全社]{ぜんしゃ:company-wide:N3}データ[基盤]{きばん:platform:N1}[刷新]{さっしん:renewal:N2}の[概要]{がいよう:overview:N1}を[紹介]{しょうかい:introduce:N2}します。\n\n#en\nRecruit Group operates dozens of services — SUUMO, Hot Pepper, Townwork, Zexy, Rikunabi, and more. Each had built its own data platform over many years, and as a result, data duplication and consistency problems became apparent. This article introduces the company-wide data platform renewal that Recruit implemented over several years.\n::\n\n::heading\n[従来]{じゅうらい:conventional:N1}[基盤]{きばん:platform:N1}の[課題]{かだい:issues:N2}\n\n#en\nIssues with the Conventional Platform\n::\n\n::para\n[従来]{じゅうらい:conventional:N1}は[各]{かく:each:N2}サービスがHadoop、Redshift、Snowflakeなどを[個別]{こべつ:individually:N2}に[採用]{さいよう:adopted:N2}しており、[同じ]{おなじ:same:N4}「[会員]{かいいん:member:N4}ID」が[異なる]{ことなる:different:N1}カラム[名]{めい:name:N5}で[管理]{かんり:managed:N2}されているケースも[多々]{たた:numerous:N4}ありました。[マーケティング]{まーけてぃんぐ:marketing}[部門]{ぶもん:department:N2}が[複数]{ふくすう:multiple:N2}サービスを[横断]{おうだん:cross:N3}して[分析]{ぶんせき:analyze:N1}する[際]{さい:when:N3}、[毎回]{まいかい:each time:N3}データの[正規化]{せいきか:normalization:N3}に[時間]{じかん:time:N5}を[要する]{ようする:require:N3}ことが[大きな]{おおきな:big:N5}[課題]{かだい:problem:N2}でした。\n\n#en\nPreviously, each service individually adopted Hadoop, Redshift, Snowflake, and so on, and there were numerous cases where the same \"member ID\" was managed under different column names. When the marketing department tried to analyze across multiple services, the time required for data normalization each time was a major problem.\n::\n\n::heading\nBigQueryへの[一本化]{いっぽんか:consolidation onto a single platform:N3}\n\n#en\nConsolidation onto BigQuery\n::\n\n::para\nリクルートは[全社]{ぜんしゃ:company-wide:N3}の[中央]{ちゅうおう:central:N2}データ[倉庫]{そうこ:warehouse:N1}としてGoogle BigQueryを[採用]{さいよう:adopted:N2}しました。[選定]{せんてい:selection:N3}[理由]{りゆう:reason:N3}には、ストレージとコンピュートが[分離]{ぶんり:separated:N1}されておりコスト[管理]{かんり:management:N2}がしやすいこと、[列]{れつ:columnar:N3}[指向]{しこう:oriented:N3}ストレージにより[大規模]{だいきぼ:large-scale:N1}[集計]{しゅうけい:aggregation:N4}が[高速]{こうそく:fast:N3}であること、そしてGoogle Cloudの[他]{た:other:N3}サービスとの[統合]{とうごう:integration:N1}が[容易]{ようい:easy:N3}であることが[挙げられ]{あげられ:cited:N1}ます。\n\n#en\nRecruit adopted Google BigQuery as the central data warehouse for the entire company. Reasons for selection include separated storage and compute making cost management easier, columnar storage enabling fast large-scale aggregation, and easy integration with other Google Cloud services.\n::\n\n::callout\n[要点]{ようてん:key point:N3}：[中央]{ちゅうおう:central:N2}[倉庫]{そうこ:warehouse:N1}に[一本化]{いっぽんか:consolidate:N3}するだけでは[不十分]{ふじゅうぶん:insufficient:N4}で、[共通]{きょうつう:common:N3}の[命名]{めいめい:naming:N3}[規約]{きやく:convention:N3}と[辞書]{じしょ:dictionary:N3}が[同時]{どうじ:simultaneously:N4}に[必要]{ひつよう:needed:N3}です。\n\n#en\nKey point: Consolidating into a central warehouse alone is insufficient — common naming conventions and a data dictionary are needed at the same time.\n::\n\n::heading\ndbtによる[変換]{へんかん:transformation:N2}[層]{そう:layer:N2}\n\n#en\nThe Transformation Layer with dbt\n::\n\n::para\nデータ[変換]{へんかん:transformation:N2}にはdbt（data build tool）を[全社]{ぜんしゃ:company-wide:N3}[標準]{ひょうじゅん:standard:N1}として[採用]{さいよう:adopted:N2}しています。dbtのモデルはGitで[管理]{かんり:managed:N2}され、プルリクエスト[経由]{けいゆ:via:N3}で[変更]{へんこう:changes:N3}が[行わ]{おこなわ:made:N5}れます。これにより、SQLが[属人化]{ぞくじんか:siloed in individuals:N1}せず、[誰]{だれ:who:N3}がいつ[何]{なに:what:N5}を[変更]{へんこう:changed:N3}したかが[追跡]{ついせき:traceable:N2}できる[体制]{たいせい:system:N3}になりました。さらに、dbtの[機能]{きのう:feature:N3}を[活用]{かつよう:leverage:N3}してテーブル[間]{かん:between:N5}の[依存]{いぞん:dependency:N2}[関係]{かんけい:relationship:N3}を[自動]{じどう:automatically:N4}で[可視化]{かしか:visualize:N1}し、データ[系譜]{けいふ:lineage:N1}を[全社]{ぜんしゃ:company-wide:N3}で[共有]{きょうゆう:share:N3}しています。\n\n#en\nFor data transformations, dbt (data build tool) is adopted as a company-wide standard. dbt models are managed in Git and changes go through pull requests. As a result, SQL is no longer siloed in individuals, and it is now possible to track who changed what and when. Furthermore, dbt's features automatically visualize dependencies between tables, and data lineage is shared across the company.\n::\n\n::heading\nデータメッシュの[原則]{げんそく:principle:N2}\n\n#en\nData Mesh Principles\n::\n\n::para\n[組織]{そしき:organization:N1}[規模]{きぼ:scale:N1}が[大きく]{おおきく:large:N5}なると、[中央]{ちゅうおう:central:N2}データ[基盤]{きばん:platform:N1}[チーム]{ちーむ:team}が[全]{ぜん:all:N3}サービスのデータを[管理]{かんり:manage:N2}するのは[困難]{こんなん:difficult:N3}になります。リクルートは[各]{かく:each:N2}サービス[領域]{りょういき:domain:N2}に「データプロダクトオーナー」を[配置]{はいち:place:N3}し、その[領域]{りょういき:domain:N2}のデータの[品質]{ひんしつ:quality:N4}と[公開]{こうかい:publication:N4}に[責任]{せきにん:responsibility:N3}を[持た]{もた:hold:N4}せています。これがいわゆるデータメッシュの[考え方]{かんがえかた:way of thinking:N4}です。[中央]{ちゅうおう:central:N2}[基盤]{きばん:platform:N1}[チーム]{ちーむ:team}は[共通]{きょうつう:common:N3}インフラと[標準]{ひょうじゅん:standards:N1}を[提供]{ていきょう:provide:N1}し、[各]{かく:each:N2}[領域]{りょういき:domain:N2}が[自律的]{じりつてき:autonomously:N2}に[運営]{うんえい:operate:N2}します。\n\n#en\nAs organizational scale grows, it becomes difficult for a central data platform team to manage data for all services. Recruit places a \"Data Product Owner\" in each service domain, who is responsible for the quality and publication of that domain's data. This is the so-called data mesh way of thinking. The central platform team provides common infrastructure and standards, while each domain operates autonomously.\n::\n\n::heading\n[個人]{こじん:personal:N2}[情報]{じょうほう:information:N3}[保護法]{ほごほう:Protection Act:N1}への[対応]{たいおう:compliance:N1}\n\n#en\nCompliance with the Personal Information Protection Act\n::\n\n::para\n[日本]{にほん:Japan:N5}の[改正]{かいせい:revised:N2}[個人]{こじん:personal:N2}[情報]{じょうほう:information:N3}[保護法]{ほごほう:Protection Act:N1}では、[利用]{りよう:use:N3}[目的]{もくてき:purpose:N4}の[特定]{とくてい:specification:N3}や[第三者]{だいさんしゃ:third party:N1}[提供]{ていきょう:provision:N1}の[同意]{どうい:consent:N4}[取得]{しゅとく:acquisition:N3}が[厳格]{げんかく:strictly:N1}に[求められ]{もとめられ:required:N3}ます。リクルートは[共通]{きょうつう:common:N3}の[同意]{どうい:consent:N4}[管理]{かんり:management:N2}[基盤]{きばん:platform:N1}を[構築]{こうちく:built:N2}し、ユーザーが[許可]{きょか:permitted:N3}した[範囲]{はんい:scope:N1}の[利用]{りよう:use:N3}しか[行えない]{おこなえない:cannot perform:N5}よう[技術的]{ぎじゅつてき:technical:N2}に[制御]{せいぎょ:control:N3}しています。BigQuery[側]{がわ:side:N3}でもrow-level securityとcolumn-level securityを[活用]{かつよう:utilize:N3}し、[必要]{ひつよう:necessary:N3}[最小]{さいしょう:minimum:N3}[限]{げん:limit:N3}のアクセス[制御]{せいぎょ:control:N3}を[実現]{じつげん:realize:N3}しています。\n\n#en\nJapan's revised Personal Information Protection Act strictly requires specification of usage purposes and acquisition of consent for third-party provision. Recruit built a common consent management platform and technically controls so that data can only be used within the scope users have permitted. On the BigQuery side as well, row-level security and column-level security are utilized to realize least-privilege access control.\n::\n\n::callout\n[補足]{ほそく:supplementary note:N2}：[個人]{こじん:personal:N2}[情報]{じょうほう:information:N3}[保護法]{ほごほう:Protection Act:N1}は[数年]{すうねん:every few years:N3}[毎]{ごと:per:N5}に[改正]{かいせい:revised:N2}されるため、[基盤]{きばん:platform:N1}[側]{がわ:side:N3}で[柔軟]{じゅうなん:flexibly:N2}に[対応]{たいおう:respond:N1}できる[設計]{せっけい:design:N2}が[望ましい]{のぞましい:desirable:N3}です。\n\n#en\nSupplementary note: Because the Personal Information Protection Act is revised every few years, a design that allows the platform side to respond flexibly is desirable.\n::\n\n::heading\n[系譜]{けいふ:lineage:N1}[追跡]{ついせき:tracking:N2}の[重要性]{じゅうようせい:importance:N3}\n\n#en\nThe Importance of Lineage Tracking\n::\n\n::para\n[全社]{ぜんしゃ:company-wide:N3}で[数万]{すうまん:tens of thousands:N3}のテーブルが[存在]{そんざい:exist:N3}する[環境]{かんきょう:environment:N1}では、「このダッシュボードの[数値]{すうち:number:N3}は[最終的]{さいしゅうてき:ultimately:N3}にどの[元]{もと:source:N4}データから[計算]{けいさん:calculated:N2}されているのか」を[追跡]{ついせき:trace:N2}することが[極めて]{きわめて:extremely:N2}[重要]{じゅうよう:important:N3}です。dbtの[系譜]{けいふ:lineage:N1}[情報]{じょうほう:information:N3}と、Dataplexなどメタデータ[管理]{かんり:management:N2}サービスを[組み合わせる]{くみあわせる:combine:N3}ことで、[影響]{えいきょう:impact:N1}[範囲]{はんい:scope:N1}[分析]{ぶんせき:analysis:N1}や[障害]{しょうがい:incident:N1}[対応]{たいおう:response:N1}が[迅速]{じんそく:quickly:N1}に[行え]{おこなえ:can be performed:N5}ます。\n\n#en\nIn an environment with tens of thousands of tables company-wide, it is extremely important to be able to trace \"from which source data is the number on this dashboard ultimately calculated.\" By combining dbt's lineage information with metadata management services like Dataplex, impact analysis and incident response can be performed quickly.\n::\n\n::heading\n[今後]{こんご:going forward:N5}の[展望]{てんぼう:outlook:N1}\n\n#en\nFuture Outlook\n::\n\n::para\nリクルートは[今後]{こんご:going forward:N5}、生成AIを[活用]{かつよう:leverage:N3}した[自然]{しぜん:natural:N3}[言語]{げんご:language:N4}でのデータ[検索]{けんさく:search:N1}、データ[品質]{ひんしつ:quality:N4}の[自動]{じどう:automated:N4}[診断]{しんだん:diagnosis:N1}、そして[国際]{こくさい:international:N3}[展開]{てんかい:expansion:N1}を[見据えた]{みすえた:envisioning:N1}マルチリージョン[対応]{たいおう:support:N1}に[投資]{とうし:invest:N3}していく[方針]{ほうしん:policy:N2}を[掲げて]{かかげて:put forward:N1}います。\n\n#en\nRecruit has put forward a policy to invest going forward in natural-language data search using generative AI, automated data quality diagnosis, and multi-region support envisioning international expansion.\n::\n",{"id":65,"title":68,"titleEn":69,"topicPath":10,"questions":70},"テック確認テスト — データ基盤とクラウド","Tech Check — Data Platforms and Cloud",[71,99,123,147,168],{"id":72,"articleId":6,"question":73,"options":76,"correctLabel":90,"explanation":93,"tags":96},"tech-data-cloud-quiz-q01",{"en":74,"jp":75},"Which is NOT cited in the article as a reason Recruit adopted BigQuery as its company-wide central data warehouse?","リクルートが全社の中央データ倉庫としてBigQueryを採用した理由として、本文に挙げられていないものはどれですか。",[77,81,85,89],{"label":78,"jp":79,"en":80},"ア","ストレージとコンピュートが分離されていてコスト管理がしやすい","Storage and compute are separated, making cost management easier",{"label":82,"jp":83,"en":84},"イ","列指向ストレージで大規模集計が高速","Columnar storage enables fast large-scale aggregation",{"label":86,"jp":87,"en":88},"ウ","Google Cloudの他サービスとの統合が容易","Easy integration with other Google Cloud services",{"label":90,"jp":91,"en":92},"エ","オンプレミス環境でも完全に同一機能で動作する","It runs with completely identical features even in on-premises environments",{"en":94,"jp":95},"The article cites cost management, columnar storage, and Google Cloud integration as selection reasons, but does not mention on-premises support.","本文ではコスト管理、列指向、Google Cloud統合の三点が選定理由として挙げられていますが、オンプレミス対応については言及されていません。",[97,98],"bigquery","data-platform",{"id":100,"articleId":6,"question":101,"options":104,"correctLabel":82,"explanation":117,"tags":120},"tech-data-cloud-quiz-q02",{"en":102,"jp":103},"What is the core of the data mesh way of thinking that Recruit adopts?","リクルートが採用しているデータメッシュの考え方の核心はどれですか。",[105,108,111,114],{"label":78,"jp":106,"en":107},"中央データ基盤チームが全サービスのデータを完全に管理する","The central data platform team fully manages data for all services",{"label":82,"jp":109,"en":110},"各サービス領域がデータプロダクトオーナーを置き自律的に運営する","Each service domain places a Data Product Owner and operates autonomously",{"label":86,"jp":112,"en":113},"全社員がBigQuery管理者権限を持つ","All employees have BigQuery administrator privileges",{"label":90,"jp":115,"en":116},"データを完全に物理的に分離する","Data is completely physically separated",{"en":118,"jp":119},"The article describes the data mesh thinking as placing a Data Product Owner in each service domain, holding them responsible for that domain's data quality and publication.","本文では、各サービス領域にデータプロダクトオーナーを配置し、その領域のデータの品質と公開に責任を持たせる仕組みがデータメッシュの考え方として説明されています。",[121,122],"data-mesh","organization",{"id":124,"articleId":6,"question":125,"options":128,"correctLabel":78,"explanation":141,"tags":144},"tech-data-cloud-quiz-q03",{"en":126,"jp":127},"Which BigQuery access control feature is utilized for compliance with the Personal Information Protection Act?","個人情報保護法への対応として、BigQuery側で活用されているアクセス制御機能はどれですか。",[129,132,135,138],{"label":78,"jp":130,"en":131},"row-level securityとcolumn-level security","row-level security and column-level security",{"label":82,"jp":133,"en":134},"ファイアウォールルールのみ","Firewall rules only",{"label":86,"jp":136,"en":137},"VPNゲートウェイ","VPN gateway",{"label":90,"jp":139,"en":140},"OS側のユーザー権限","OS-level user permissions",{"en":142,"jp":143},"The article explicitly states that, in addition to a common consent management platform, row-level and column-level security on the BigQuery side are utilized to achieve least-privilege access control.","本文では、共通同意管理基盤に加えBigQuery側でrow-level securityとcolumn-level securityを活用し、必要最小限のアクセス制御を実現していると明示されています。",[145,146],"privacy","security",{"id":148,"articleId":149,"question":150,"options":153,"correctLabel":86,"explanation":162,"tags":165},"tech-data-cloud-quiz-q04","tech-digital-cho-cloud",{"en":151,"jp":152},"Which certification system is mandatory for entering the Government Cloud?","ガバメントクラウドへ参入するために必須とされている認証制度はどれですか。",[154,156,158,160],{"label":78,"jp":155,"en":155},"ISO 9001",{"label":82,"jp":157,"en":157},"PCI DSS",{"label":86,"jp":159,"en":159},"ISMAP",{"label":90,"jp":161,"en":161},"SOC 1",{"en":163,"jp":164},"The article explicitly states that ISMAP (Information system Security Management and Assessment Program for government information systems) certification is mandatory to enter the Government Cloud.","本文ではガバメントクラウドへ参入するにはISMAP(政府情報システムのためのセキュリティ評価制度)の認証が必須と明示されています。",[166,167],"ismap","certification",{"id":169,"articleId":149,"question":170,"options":173,"correctLabel":82,"explanation":186,"tags":189},"tech-data-cloud-quiz-q05",{"en":171,"jp":172},"According to the article, what is the main reason the domestic cloud Sakura Internet is included in the Government Cloud?","国産クラウドのさくらインターネットがガバメントクラウドに含まれている主な理由として、本文で示唆されているものはどれですか。",[174,177,180,183],{"label":78,"jp":175,"en":176},"海外クラウドより常に安価だから","It is always cheaper than overseas clouds",{"label":82,"jp":178,"en":179},"経済安全保障および地政学リスクの観点からの国産クラウドの育成","Cultivating domestic clouds from the perspective of economic security and geopolitical risk",{"label":86,"jp":181,"en":182},"技術力が他社より圧倒的に優れているから","Its technology is overwhelmingly superior to others",{"label":90,"jp":184,"en":185},"全自治体がさくらの株主だから","All municipalities are Sakura shareholders",{"en":187,"jp":188},"The article notes as a supplementary point that depending solely on overseas clouds carries large geopolitical risk, so cultivating domestic clouds is included as a policy goal.","本文では、海外クラウドのみへの依存は地政学リスクが大きいため、国産クラウドの育成が政策目的として含まれていると補足されています。",[190,191],"government-cloud","economic-security"]