How companies introduce AI in a structured way: from strategy to implementation

Artificial intelligence (AI) is revolutionising companies by optimising processes, reducing costs and opening up new business opportunities. According to a McKinsey study (2022), the potential is clear: in marketing, product and service, revenue growth is especially possible, while supply chain and operations offer cost reduction. But how do you implement AI in a structured and sustainable way?

The foundation: the AI Strategy Canvas

To use AI effectively, companies must define their strategy clearly. We use the AI Strategy Canvas, which is divided into 9 key questions:

CreativeCommons BY 4.0, 2025 Mike Schwede

1. Vision & values:
   Why do we use AI? How does it fit with our company values?

2. Business objectives:
   Which company objectives does AI support?

3. Challenges / pain points:
   Which processes or tasks cause high costs or take too much time?

4. Relevant data & technologies:
   Which data sources and platforms are available? Which AI technologies make sense?

5. Use cases & prioritisation:
   Which 3-5 AI use cases deliver the greatest value?

6. Implementation & integration:
   How do we embed AI into existing processes and systems?

7. Organisation & roles:
   Which departments and experts are involved?

8. Ethics, data protection & governance:
   What risks are there? How do we ensure data protection & compliance?

9. Roadmap & next steps:
   Which milestones do we set for the next 12-24 months?

Step by step to AI integration

After strategy comes implementation. This happens in 3 stages, from first tests to full automation:

Stage 1: Automate individual tasks

• Use of ChatGPT & LLMs for simple tasks

• Test first prototypes, e.g. AI-generated content or analyses

• Focus on quick wins

Stage 2: (Partial) automation with your own data

• Use of standard tools (e.g. CustomGPT, AI video generators)

• Initial data integration & workflow optimisation

• Process analysis: Which steps should be fully automated?

Stage 3: Full automation

• Integration of AI models into existing IT systems

• Train your own AI models with internal data

• Implement scaling & governance mechanisms

Next steps: the 3 most important initial measures

1️⃣ Develop AI principles & policies

📌 Define clear guidelines for the use of AI in the company
📌 Establish ethical guidelines and a data protection concept

2️⃣ Train employees & management

📌 Offer AI workshops & training for teams
📌 Prepare managers for strategic use of AI

3️⃣ Develop the AI Strategy Canvas

📌 Define a roadmap for the use of AI in interactive workshops
📌 Prioritise use cases and start initial tests