Maximising ROI on AI: 6 Dos and Don'ts for Business Leaders Organisations with a balanced, forward-thinking vision will outperform those that chase the early hype.
By John Atkinson Edited by Jason Fell
Opinions expressed by Entrepreneur contributors are their own.
You're reading Entrepreneur Europe, an international franchise of Entrepreneur Media.
![](https://assets.entrepreneur.com/content/3x2/2000/1737478555-ai-roi-dos-and-donts-leaders-0125-g1979289147.jpg)
As artificial intelligence (AI) adoption spreads across the enterprise world, the gap between hype and measurable results grows wider. Business leaders are navigating a whirlwind of AI-driven possibilities, but too many are investing without a clear roadmap for success, leading to underwhelming returns. To fix this, optimism must be met with pragmatism; enthusiasm tempered with realism, which is why some clear practical advice to help avoid common pitfalls and create value is needed.
Those who get it right will be rewarded: last year, it was reported that a third of European businesses have adopted AI – a 32% growth rate since the previous year. If the momentum continues, by 2030, this could contribute a staggering €600 billion in gross value added (GVA) to the European economy – equivalent to the entire European construction industry.
Here are some key dos and don'ts to help maximise ROI on AI:
Do: Put end-user experience first
Ultimately, the purpose of AI should be to enhance the experiences of customers and employees. From automated chatbots to data-driven analysis, the goal is to make their digital touchpoints more seamless, personalised and meaningful.
This is where AIOps and unified observability solutions can really help. Unified observability connects all data points across an organisation into a single view, standardising alerts while simplifying analysis and prioritisation. AIOps detects anomalies across platforms, generates smart recommendations, and automates tasks. Together, these technologies allow you to proactively monitor the performance of IT estates.
This helps prevent downtime, boosting productivity while reducing stress. Taking this approach also ensures that the deployment of AI is successful and that the needs of all end-users are intuitively supported. It's all about making sure the technology works faultlessly for those on the front line. Think user-first, and AI can empower the people directly responsible for driving commercial growth.
Do: Think long-term
AI is not a quick fix. Its transformative benefits often take time to materialise, demanding patience and adaptability. You should therefore approach it as a strategic investment in sustainable growth, rather than a shortcut to resolving operational challenges.
For example, retailer and manufacturer Zara successfully uses AI to track customer behaviour, manage its inventory and build a more resilient supply chain; each step compounding the improvements of the one before. It holistically considers the problems of tomorrow, while still meeting market demands today.
The takeaway here is to establish a strategy that encompasses all operational components — and create a clear, phased roadmap that all stakeholders can follow to avoid focusing on the short-termism.
Do: Empower cross-generational teams
Generational attitudes toward AI can vary substantially, which can hinder adoption if not handled carefully. It's understandable that older generations – specifically Baby Boomers and Gen X – are less comfortable with AI in the workplace than their younger Millennial and Gen Z counterparts, who have grown up as digital natives.
But without accommodating the views of all employees, enterprise-wide adoption will always be impeded, which means losing out not only on AI's benefits but also the collective expertise of a company's human talent. Dedicated AI teams that reflect the full diversity of your workforce — generationally and otherwise — are more likely to embrace the changes and help identify blind spots in the process.
As well as fulfilling the ethical duty to create inclusive environments, cross-generational collaboration like this invites a broader range of perspectives that combines the tech savviness of digital-native generations with the strategic experience of seasoned professionals.
Don't: Overestimate your readiness
AI certainly has the power to elevate performance, but many organisations misjudge their readiness, leading to costly missteps. Conducting a realistic self-assessment of your capabilities — including areas such as data quality, digital infrastructure capacity, team expertise and even workplace willingness — can help prevent this and keep ambitions grounded in achievable outcomes.
Similarly, benchmarking against competitors or sector leaders can expose gaps and highlight opportunities. For instance, logistics giant DHL routinely evaluates its supply chain capabilities to identify areas where AI can optimise route planning and accurately forecast demand.
By doing this, DHL ensures that the integration of AI delivers tangible benefits without disrupting workflow. And if the reflective process reveals an unreadiness or inefficiency, then it can be swiftly addressed before any further budget is allocated towards it.
Don't: Ignore your network infrastructure
Seamless data flow is the lifeblood of AI systems. Even the most advanced algorithms will fail to deliver value if the underlying network isn't optimised for AI's demands. Latency issues, unstable connectivity or inadequate storage capacity can derail even the most promising AI strategies.
With the support of solutions that provide data-driven insights, your business can gain a more comprehensive understanding of its network's strengths and weaknesses.
Armed with that knowledge, IT teams can then make real-time decisions that augment the efficiency and reliability of their applications. In turn, this makes it easier to prevent the kinds of data bottlenecks that destabilise an AI initiative before it delivers true value.
Don't: Overlook hidden blind spots
Even though AI depends on high-quality data, organisations commonly overlook key data sources that can severely hinder AI performance if left siloed. Shadow IT, corporate mobile devices and even zero-trust environments are just some of the areas that can create blind spots.
Within the financial industry, for example, it can happen when outdated digital architecture renders valuable data inaccessible, prompting banks to invest more in technology modernisation. However, if siloed datasets can be consolidated, organisations can reduce inefficiencies and generate more complete data. Crucially, they can also prevent overspend.
Conducting a thorough audit of data sources is the only way to ensure no stone is left unturned during the pursuit of digital transparency. Rectifying blind spots upfront mitigates the risks associated with AI integration — paving the way to better outcomes.
AI is a truly powerful tool — but how transformative it will be ultimately hinges on how you use it. A foundation for sustained success can be built by exercising patience, prioritising user impact and unifying network infrastructure. Likewise, being realistic with expectations, addressing blind spots and curating diverse teams will also help with avoiding common pitfalls.
Logically, the organisations with a more balanced, forward-thinking vision will outperform those that chase the early hype. To make AI worth its weight in gold, pair ambition with pragmatism; turning AI potential into performance, and investment into returns.