By Yemi Olagbaiye, director, client portfolio at digital software consultancy Softwire
2024 is a landmark moment in the enterprise world in terms of putting its money where its mouth is with generative AI. According to McKinsey, 65% of businesses are now regularly using genAI as companies around the globe gambit on its game changing potential.
This surge of experimentation is backed by serious levels of investment. New research from Boston Consulting Group estimates that genAI will increase to nearly 5% of corporate IT budgets this year, with that figure set to grow 60% over the next three years.
But like all new tech, a willingness to embrace and invest does not guarantee success. In a sobering assessment of the market to date, the Harvard Business Review reports that as many as 80% projects involving genAI currently fail to get off the ground – despite its infinite hype.
There are many reasons why such a promising tool is unable to generate value, not least the immature and experimental nature of the tech itself. But from my experience at the front end of the genAI wave, one explanation stands out: unrealistic expectations.
Levi’s, Instacart and Samsung thought they were getting a jump on their rivals when they took on the race for genAI maturity. But all three brands – and many others – have suffered a backlash due to taking on too much, too fast.
Before investing, C-suites must understand the nature of the genAI opportunity they’re mulling over and what is at stake. Rather than default to the best headline idea, they also need to choose a level of investment that realistically fits their budget, talent and immediate growth plans. To aid that process, here are the three key tiers of AI investment for leaders to consider:
- Quick-win AI investments: Quick wins are accessible AI investments that can be introduced by most organisations, irrespective of scale or specialism. They are typically productivity assistants or task-specific AI tools that are rapidly deployed to enhance efficiency in day-to-day operations.
Examples of genAI-powered tools that can deliver quick wins include Microsoft Prompt Flow, a development tool that simplifies experimenting, prototyping, iterating and deploying AI applications. Meanwhile, coding assistants like Codiga and Tabnine help developers write code more efficiently.
The no-brainer advantages of quick-win projects are that they are low-cost, low-risk, easy to integrate, and the time to ROI is relatively quick. On the flip side, however, these fast-delivery initiatives will increasingly struggle to achieve differentiation over time, especially in terms of productivity gains.
- Competitive AI investments: This is where genAI starts to get interesting as the stakes climb steadily higher. It involves ambitious ventures that embed AI into domain-specific or custom applications. The goal is to amplify core business processes and create a lasting competitive advantage in any market.
Examples of this genAI investment include Tax Systems’ compliance co-pilot Alphamap, which takes over data inputting and prepares initial tax treatments. Delivering a time reduction of up to 80%, the co-pilot frees tax professionals for more critical tasks such as business partnering and data insights.
This type of project is a good bet if you want to deliver better, more sustainable genAI outcomes over a longer timeframe (driving value over, say, one to two years). But it also requires more patience, budget, and an ability to hold your nerve while the changes you’re aiming for take effect.
This category typically involves significant platform investments with moderate to high costs. It tends to be more complex in terms of tech integration and may require some level of talent recruitment. Ongoing monitoring is also a crucial part of the differentiation strategy to avoid escalating costs and the risk of scope creep.
- Transformative AI investments: These are high-stakes, “holy grail” initiatives that have the potential to create whole new products, markets, or business models. When companies think about genAI, they typically have this in mind. Though for many, heading here first is a bit like choosing a Porsche as your first car; in other words, it may be a reach too far.
Transformative AI platforms include BloombergGTP, which was purpose-built to improve existing natural language processing tasks such as sentiment analysis, named entity recognition, news classification, and question answering. Using a blend of Bloomberg’s financial data archives and public datasets, the platform gives financial professionals quick and accurate insights to inform and accelerate decision-making.
While transformative AI has the biggest impact in terms of its ability to disrupt markets and create new business models, it also comes with the highest costs—alongside a miasma of technical, financial, and market risks. What’s more, the time-to-value arc is usually over two years, which requires confident and committed leadership.
There’s little doubt that successful genAI integration requires relevant technical expertise – and this is usually what dominates the discussion when boards are scoping out an initiative. But it’s worth noting that gauging where your organisation sits on the genAI spectrum is the most important first step – and it’s also more of a leadership issue.
To thrive, every enterprise activity involving genAI will need to be located somewhere within the above spectrum of options. Deciding where and whether there is capacity for progression demands careful consideration. It also depends on a range of overlapping factors—such as industry context, organisational readiness, the quality of data at your disposal, strategic goals, and the potential for value creation. Navigate these questions successfully, and genAI will catalyse tangible, tailored growth that every business hopes for.
