Davenport and Mittal’s research (Davenport & Mittal, 2003) identified ten actions essential to successful AI adoptions for enterprises. It argued that AI challenges the fundamental way humans and machines interact with each other and, therefore, should and will change all aspects of the business.
The ten undertakings are summarised below:
Know what you want to accomplish
Improved productivity, better products, cost reduction, or better marketing outcomes? A guiding principle is a prerequisite.
Work with an ecosystem of partners
Buy or build? It is often too expensive to build all the necessary capabilities in-house. You need to identify the right partners to work together.
Master Analytics
Companies should be committed to using data and analytics for decision-making. What is our unique or proprietary data? Is it of excellent quality? Readily available data only results in similar machine-learning models and outcomes as everyone else. Notable examples include Segata Technology’s automated visual inspection of silicon wafers for hard drives.
Create a modular, flexible IT architecture
Like any IT project or digital transformation, the biggest headache is not the lack of new technologies or tools; it’s the complicated legacy systems that always seemed impossible to deprecate. AI adoption is no different. To incorporate any emerging tech into the existing infrastructure to realise the value of integration, a flexible and lean information architecture is the runway to success.
Integrate AI into existing workflows
Start by determining which workflows are ripe for AI speed and intelligence. Begin integrating AI into them as soon as possible.
Build solutions across the organisation
To get the maximum benefit from AI, you’ll want to deploy it throughout the organisation. Designing one algorithmic model for one process will not give you a unified approach that can be replicated across the company.
Create an AI governance and leadership structure
The best leaders know what AI can do in general, what it can do for their companies, and what implications it might have for strategies, business models, processes and people. The biggest challenge is creating a culture that encourages data-driven decisions and excites employees about AI’s potential.
Develop and staff centres of excellence
Companies need considerable talent and training in AI, data engineering, and data science to adopt AI successfully.
Invest continuously
Tech capacity is an expensive living creature. It needs constant maintenance and investment, especially if a firm wants it to have a big impact on its bottom line. The success of any AI strategy will require more than a one-off investment.
Always seek new sources of data
Data is the deciding factor that makes a difference in AI performance. To differentiate and build the AI capacity that strengthens the firm’s competitive advantage, firms must build robust data pipelines that allow them to maximise the value extraction of their most valuable assets in the digital age - the high quality data.
References
- Davenport, T.H. and Mittal, N. (2023). Stop Tinkering with AI. [online] Harvard Business Review. Available at: https://hbr.org/2023/01/stop-tinkering-with-ai.