Reading time: 2 & 10 minutes

Hi, and welcome to the second part of my deep dive into the state of AI. I explore Real usage, Challenges, and Winning Strategies using real data from surveys totaling 3000 companies.

⚡ Your 2-Minute Summary

In Part 1 (link), we uncovered how businesses use AI and where they see value, finding that only 26% of companies generate tangible returns despite massive investments. This second part dives deep into why the majority struggle and what differentiates successful implementations from failures.

I've combined my experience with enterprise survey results from more than 3000 companies (conducted by BCG, IBM, Deloitte, McKinsey, Dataiku, and more).

Key Findings:

• Most companies remain stuck: 70-80% of AI initiatives never move beyond the pilot phase. However, most of the value is gained when AI is scaled with 3x ROI and a 30% premium on key financial metrics (Accenture Report: Built to Scale).

• Organizations must address three critical challenge areas, in order of priority: People/Process (70%), Technology (20%), Algorithms (10%)

• Success markers: Companies that scale successfully, invest more, plan, and execute a comprehensive enterprise-level/multi-year strategy.

This 70-20-10 split challenges a common misconception: while many organizations focus on technical capabilities, successful AI implementation is predominantly a people and process challenge. Organizations that recognize and act on this insight tend to achieve better outcomes. AI must be treated as a holistic approach, not a confined IT technology. It might be as disruptive as the internet.

Challenges and Barriers to Scaling AI

Despite considerable investments in 2023-2024, Deloitte's research reveals that 70% of organizations have moved less than 30% of their AI experiments into production (Deloitte Q3 2024, p.27).

Measuring & Proving AI's ROI potential is more challenging than technical limitations:

Looking at the data below (BCG), the top three barriers are all related to business value realization rather than technical limitations:

• 66% struggle with establishing ROI on identified opportunities

• 59% face difficulty prioritizing opportunities against other concerns

• 56% struggle with making a business case for scaling initiatives

This suggests that the primary challenge isn't technical capability but demonstrating and capturing business value.

Lack of organizational level strategy with people at the heart of implementation:

Data & IT infrastructure readiness constitute a barrier to entry for many organizations. However, the lack of an overall strategy and change management keeps most companies from the trial phase. These similar patterns emerged when I examined BCG, Dataiku, Deloitte, and Accenture's report.

According to Dataiku's report, 21% do not specifically measure ROl on Generative A, while 64% do not consistently measure the value from AI. If ROI is not tracked, there will be no concrete evidence to justify the continuous or increase of investments (beyond just hypotheses). BCG's report confirms this assumption, with 66% of respondents mentioning the difficulty in establishing ROI on identified opportunities.

Once the strategy is set and the ROI metrics are established. The next biggest challenge is AI talent. With the surge of AI adoption, organizations worldwide struggle to hire AI engineers or upskill people in AI literacy.

• 37% require AI engineers

• 28% struggle with talent acquisition (Source: BCG Report)

Data & Technology Integration:

40% of companies globally say that data-related challenges are slowing GenAI efforts globally, with 55% of organizations avoiding certain AI use cases due to data issues (Deloitte). Most of these data challenges relate to the following (IBM AI in Action report):

• Lack of access to high-quality data: Inaccurate data and bias

• Data security

• Regulatory changes around data privacy and confidentiality

Interestingly, most of these challenges are also concerns for these organizations.

Most organizations want to adopt AI and benefit form it’s technological progress. However, a proper data strategy is crucial to ensure result accuracy and optimal benefits.

Integration Complexity with existing systems:

Integration with existing systems, cited by 56% of respondents, represents a unique challenge because:

• It spans organizational boundaries

• Requires coordination between multiple stakeholders

• Impacts existing workflows and processes

• Often requires significant change management

As AI impacts the whole organization, working with it implies a holistic integration of AI within existing IT systems. For now, most organizations don't know how to do that effectively.
Moreover, according to Dataiku, 48% struggle with limited AI adoption budgets, with 57% using their existing IT budget for organization-wide AI implementation. This creates a tension between maintaining current systems and innovating for the future.

What successful companies do differently

Let's analyze the benchmark between AI leaders and companies that are still in the process of adopting AI, referred to as AI learners. We find that AI leaders implement a comprehensive and well-thought-out strategy, invest more resources, and primarily concentrate their efforts on people and processes.

Leaders in AI tend to invest more and focus on generating revenue to a greater extent than their competitors:

During a recent survey by IBM & Dataiku, key differentiators in investment are:

• 71% aggressively invest in AI (vs 19% for AI adopters)

• 54% plan to spend over $1M on GenAI in next 12 months

• Focus on revenue generation over cost reduction (36% vs 26% for non-leaders)

This investment pattern suggests successful organizations view AI as a strategic capability rather than a tactical tool. The higher investment levels correlate with more comprehensive implementation strategies and better business outcomes.

AI leaders widely adopt artificial intelligence but focus intensively on high-return-on-investment (ROI) use cases.

According to IBM (AI in Action 2024), leaders are approximately 80% more likely than Learners to invest in the top 4 use cases:

• Customer experience

• IT operations & automation

• Virtual assistants

• Cybersecurity

This difference is further confirmed by Dataiku's report on the difference in use cases between AI leaders (Defined as pioneers here) and non-leaders:

This difference ties back to the lack of understanding of AI capabilities and where it can bring the most value. It might be that organizations converge to these use cases after multiple pilots across business lines, or it might be that their strategic department or third-party consultants strategically decided to use AI based on where it is already providing value.

Leaders focus on core business functions:

As we've seen in Part 1, most of the current value is achieved in core business functions (70-90% of value).

Leaders direct their efforts toward people and processes more than technology & algorithms:

Since most challenges originate from people & processes, it's natural that the effort must be geared proportionally. More precisely, BCG gives a detailed breakdown of focus areas by segment, with 70% on people & processes, 20 % on technology, and 10% on algorithms. We tend to focus on technology. However, key drivers (& barriers) are usually empowered by people & processes more than anything else.

These findings are confirmed by another study from IBM among IT professionals where we see the largest discrepancies between leaders in AI and others in limited AI expertise or knowledge coupled with a holistic AI strategy are the biggest outliers.

Another study from IBM confirms the importance of alignment with 72% of AI leaders saying that their C-suite is fully aligned with IT leadership about what needs to be done to achieve AI maturity. This number drops to 36% for AI learners.

Looking Ahead

In Part 3, we'll provide a detailed playbook for successful AI implementation, building on these insights to create a practical framework for business leaders. I'll explore specific strategies for overcoming the challenges identified.

Sources

• Deloitte State of AI Q3 2024

• BCG Value in AI Report 2024

• Dataiku Survey : AI, Today

• Accenture: Built to Scale Report

• IBM Global AI Adoption Index Report

📧 Was this email forwarded to you? Subscribe below

If you liked this first part, consider subscribing to and sharing the newsletter. It will help me do what I like: research and write about topics related to tech and business.

Sources:

• Deloitte State of AI Q3 2024

• BCG Value in AI Report 2024

• Dataiku Survey : AI, Today

• Accenture: Built to Scale Report

• IBM Global AI Adoption Index Report

Amine Rabehi

Find me on LinkedIn or Book a Free Call

🔍 Whenever you’re ready, here are 3 things I can help you with:

• Technology Solution/Product consultancy: I help you assess your product idea from a technological & business point of view. So far, I have built a network of professionals covering the MENA region in term of IP, Market Research & Complex AI solution development.

• Market entry strategy and support: understand the market, legal framework

• Tech Product Development Solutions: I’ve managed more than 15 tech product development projects (hardware & software). I help you go from idea to initial batch launch.