Almost every headline, vendor pitch, and business strategy conversation is about AI. You don't imagine it. All business leaders are under pressure to do something with AI.
However, most organizations struggle to gain traction and turn pilots into real business impacts. a Boston Consulting Group Report Since October 2024, we have found that 74% of companies have shown tangible value from AI and are struggling to move beyond proof of concept. Recent surveys by MIT This trend suggests that it will continue in 2025. We look firsthand at the same challenges as the companies we work with each day.
The problem is not technology. This means that most organizations do not build the preparation or capacity to use effectively.
Over the past few years, we have begun our AI journey with a variety of businesses in Iowa and beyond. Below are five principles we deemed most important to an organization's success with AI.
1. Solve real problems, not trendy problems
It's tempting to follow what others in the industry are doing, but AI isn't perfect for all sizes. You need to adapt your specific workflows, data, and goals. They ask things like, “Where are we bottlenecked?” “What slows down our team?” or “Are you lacking insights that can improve your outcome?”
The best impact comes from eliminating operational bottlenecks. This is something we consistently see in our work. Target the points that relieve pressure to unlock time and efficiency.
2. Start with the data you already have
You don't need “perfect data” to start using AI effectively. Clean and structured data is important for certain types of AI tasks, such as unsupervised machine learning and prediction.
By starting clean data quickly rather than worrying, you learn which gaps are really important and where targeted data cleanup is worth more investment and time.
3. Balance of responsibility and efficiency
AI works best when increasing human potential, not when blindly replacing it. It can process data quickly, reduce cognitive load and eliminate repetitive tasks, but it doesn't understand the ethical implications of customer relationships, team dynamics, or its output.
Make sure you know the balance between responsibility and efficiency by asking important questions like, “What is the result of exposing this data to AI?” “Are we making our teams easier, or are we quietly burdening them with more uncertainty?” or “What are the potential risks to our business and others?”
The best use of AI will strengthen your teams, keeping people in a loop and keeping them transparent with customers and partners.
4. Use proven tools before building custom
Most organizations don't need to train their own models to see the real impact within their business.
Organizations like Openai, Anthropic, Google, and other organizations invest heavily in improving LLM every day, allowing them to build their progress directly.
Many problems have already been solved, and existing AI tools can often adapt quickly. Take Language: Your organization may use technical terms or acronyms that popular AI does not understand. Instead of building custom models, use approaches such as Getted Generation (RAG) to efficiently inject business contexts.
5. Please don't go alone
Like other tools, AI has a learning curve. Many companies can make progress on their own, but often it takes more time and involves more trial and error.
Whether you're a small company with no deep technical resources or a large company with complex systems, the lessons are the same. Don't try to learn everything in the difficult way. Working with partners, peers, or industry groups will help you learn faster, avoid common failures, and see value faster.
Conclusion
AI can absolutely move needles for business. By focusing on real problems, starting with the data you have, balancing responsibility and efficiency, using proven tools, and learning with others, you're much more likely to turn your AI pilot into real business value.
Christina Colson is the AI strategy lead for Lean Technique. Lean Techniques work with businesses of all sizes to make technology a competitive advantage. If you are on your AI journey and can use an external perspective, then Leet's Connect: LeanTechniques.com
