Nissin Foods' AI strategy: CIO Narita on company-wide rollout
【日清食品のAI戦略】ChatGPT導入で組織はどう変わる？CIO成田氏が語る全社浸透のリアル
B · Useful, limited First-hand account from Nissin Foods group CIO Toshihiro Narita of a company-wide generative-AI rollout with a dated adoption curve and named interventions; self-reported figures, probably recorded in 2025.
Summary based on the full transcript·Original: about 46 minutes to watch
In short
An episode of GLOBIS's AI Business Shift series for managers (published 15 October 2026) with Toshihiro Narita, executive officer and group CIO of Nissin Foods Holdings, interviewed by Koji Torikata of GLOBIS. Narita describes rolling out an internal GPT environment to about 5,000 white-collar staff from April 2023. Company-wide prompt training did not lift daily use from 3-5%; shared prompt templates did, followed by a sales-first department push and mandatory manager training. Narita says the rollout started "two years ago", so the figures probably date from 2025.
Key numbers
- ~5,000white-collar staff across the Nissin group with access to the internal GPT environment
- 3-5%daily utilisation in the first weeks after launch in April 2023
- 60-70%company-wide utilisation at the time of recording
- 100%utilisation reached in the internal audit department
As reported by the source.
Japanese·Video·opens youtube.com
What it reports
- The environment went live in April 2023, about six weeks after GPT-4. The trigger was a demo to top management in which the CEO asked for mobile access on the spot.
- Daily use in the first weeks was 3-5%. People did not want to write long prompts, and prompt-engineering training did not change the number.
- Use rose after the team built shared prompt templates that staff could call and lightly edit. For sales, each salesperson listed candidate uses, templates were refined for about a month and then shared with all sales staff.
- Rollout went department by department, starting with sales because its head wanted the result, then marketing, corporate planning, production and audit. Each success was publicised quickly.
- Internal audit reached 100% utilisation after one AI-literate auditor changed his own workflow and the team copied him.
- A ranking of heavy users by name in the admin console identifies champions, who then teach in the internal training programme. The delivery team is about 13 volunteers from IT.
- Managers take mandatory training. Narita says a manager who does not push the team to use AI lacks a management skill, and that working without AI is no longer acceptable.
How they got it
Interview. Figures come from Narita's own reporting and admin-console data; no method or definition of "utilisation" is given beyond daily use.
Limits
- Recording date is not stated; the 60-70% figure is probably from 2025.
- Self-reported, with no outcome measures beyond usage.
- Part of a GLOBIS training series, so framed for managers rather than for analysis.
Why it matters for companies
The finding that prompt training did nothing while shared templates moved use is cheap to act on. So is the sequence: start with a department whose head wants the result, publish wins fast, find champions from usage data, and make managers accountable. It also shows a non-tech manufacturer reaching high everyday use without a large central team, though usage is not the same as business results.
Key passages in translation
Japanese original
日時で見た時の利用率3%とか5%でもう本当不思議すぎてなんでみんなこれ使わないのかな
Daily utilisation was 3% or 5%, and it was really puzzling: why was nobody using this?
Japanese original
1ヶ月間にあのブラッシュアップをしたプロンプトのテンプレートを作り、それを全営業担当者で共有する
We made prompt templates refined over a month and shared them with all sales staff.
Japanese original
もうそもそもAIを使わずに業務をやるっていったことはもう許容されないと思ってます
I think doing your work without using AI is simply no longer acceptable.
Summary based on the full transcript. Read the original for the full argument and data.Watch the original on youtube.com
- Tags
- AI enablement, AI adoption