Since the launch of ChatGPT at the end of last year, it seems that the world can’t stop talking about AI, its potential and our fears surrounding it. According to recent research, 35% of companies are now using AI and a further 42% are exploring its potential. This willingness to embrace new technologies, especially by larger enterprises is encouraging and is undoubtedly helping to increase innovation.
However, many are jumping in headfirst, investing significantly in AI without much consideration for the use case and how it will drive value. Over two thirds (76%) of executives face challenges when scaling the AI implementations and, as a result, a whopping 85% of data, analytics and AI projects fail. Unfortunately, this means many organizations will fail to attain the transformation results they seek. With UK productivity at all time low, most organizations cannot afford to be wasting much-needed resources this way.
Get your data foundations right
Before even considering potential use cases, organizations need to assess the issues their data may create. Poor data quality, infrastructure, transparency and accessibility are the most common challenges. An organization-wide analysis of teams, and how each uses and accesses data, can help evaluate capabilities. At the same time, adopting a decentralized and democratized approach, such as a data mesh model, will ensure data is owned and maintained by those who understand it and can provide access to those that need it.
For an organisation to reap the rewards of AI, it often must change the way it approaches data first. Considering data as more of a product, rather than a technical asset, is a step in the right direction. Data must be deeply integrated into an organisation’s core through a sustainable and scalable strategy that is embraced across the organisation.
Michael Chalmers is CEO of Mesh-AI.
Five pillars of an AI strategy
An AI strategy can begin once an organization has established strong data foundations. Once that has been established, there are five core pillars of an effective AI strategy that covers people, processes and technology:
The first, and arguably most important, is culture. Creating a culture where individuals at all levels are well-informed about potential applications and business value, aligned to the business goals, will impact the success of AI significantly.
The second is ideation. Before putting resources into any AI solution, organizations should work with data experts to test if ideas are both valuable and technologically feasible. Without ideating effectively, organizations will fail to recognize major opportunities, and any AI initiatives developed are likely to be sub-optimal.
Next is delivery. Once the most beneficial use cases are identified, in line with business objectives, strong policies need to be in place to ensure end-to-end delivery. This includes setting-up, training, measurement, scaling and maintaining AI models over time. Without an effective framework, and maintenance in particular, AI initiatives will never evolve to reach the volume and scale required to make an impact.
The fourth pillar is trust. A lack of trust has the potential to be the biggest barrier to AI adoption. A recent study on trust showed that nearly half of all employees are wary of trusting AI in the workplace. If AI is to be widely accepted, its vital that the solutions are clear, ethical and abide by regulations. This will help employees see that the technology is safe, transparent and accountable, building trust.
Finally, the impact on the organization. AI initiatives should never be launched without a clear business goal. It is vital for any organization adopting AI to set clear performance metrics and track return on investment at the start to measure its impact – as with any significant investment.
Reaping the business benefits
AI has the potential to add business value in a variety of ways from improving productivity to operational efficiencies, cost savings and helping organizations to automate tasks, decision making and to innovate. However, many embark on projects in isolation without considering the potential on wider operations.
Investing the time to develop a deep-rooted AI strategy centred around building a data-based culture with solid ideation, delivery, trust and impact can reap deep benefits. Those organisations that take this approach can drive innovation, growth and outcomes that will help gain a competitive advantage in a rapidly evolving technology landscape.
