Have you ever found yourself doing something because everyone else is doing it? Have you ever felt like you have to do something because if you don’t, you’ll fall behind? This is exactly what’s happening with GenAI adoption in businesses today. Companies are rushing to implement artificial intelligence, often without a clear strategy, driven by FOMO and market pressure.
In this article, you’ll discover why this “bandwagon effect” is leading to AI implementation failures and learn 5 practical tips to ensure your organization’s AI initiatives succeed in 2025.
Whether you’re considering AI adoption or already struggling with implementation, these insights will help you move beyond the hype and create real business value.
What Does AI Have to Do with the Bandwagon Effect?
The bandwagon effect is a psychological phenomenon in which people do something primarily because other people are doing it, regardless of their own beliefs. This effect is particularly visible in the IT industry, especially in the field of AI. The widespread enthusiasm around Generative AI (Gen AI) has intensified this trend. This is concerning because people have started to think of AI as a cure-all. AI can undoubtedly assist humans in many tasks.
Why Do Most Companies Fail with AI Implementation?
Many businesses are investing time and money in AI simply because it’s trendy, rather than because it addresses specific business needs.
Furthermore, many decision-makers who implemented AI are disappointed with the results. The return on investment has fallen short of expectations. A study conducted by Deloitte shows that “there was a 29% increase in the number of respondents self-identifying as ‘underachievers,’ suggesting that many organizations are struggling to achieve meaningful AI outcomes.“
Consequently, organizations become discouraged and reluctant to continue investing in this promising and evolving technology. As a result, their businesses remain tied to traditional models, becoming less competitive and resistant to change.
How to Increase Your GenAI Implementation Success Rate
There are many reasons why companies fail to implement AI in their business. In most cases, it’s not just one simple reason—it’s usually a combination of different factors that should be considered holistically. In this article, I’ll share 5 simple tips that can help guide your AI journey and increase your probability of success.
Note that this isn’t a universal solution. These observations come from my experience working closely with clients and managing various AI-related projects. Take what best fits your situation.
1. Start with the Problem
As with any project, it makes no sense to start if you don’t know what kind of problem you want to solve, what challenges your company faces, or what processes you want to improve. Simply put, start with the business case. Don’t think about technology—technology is just a tool, a means to solve your problem.
Many people make the mistake of starting with the tool rather than the problem. Many of our customers come to us saying “I want AI,” but when we ask why they need it, they struggle to answer this question.
Before you start thinking about technology, take time to consider these essential questions:
- What problem am I trying to solve?
- What challenges does my business face?
- What processes can be automated?
- Why should my company invest money in this solution?
- Is it costly?
- Is it time-consuming?
- Is it error-prone?
These questions may seem simple, but they often prove the most challenging and time-consuming to answer. Don’t underestimate them. In our “as soon as possible” culture, give yourself time to think rather than rushing to act. You’ll quickly discover the benefits of this approach. Remember that according to the DSDM philosophy, “best business value emerges when projects are aligned to clear business goals.” Think strategically and ensure only the right projects move forward.
And no, AI itself isn’t a business case.
2. Consider Limitations and Special Requirements
Every problem has a context – the surrounding factors that determine a project’s success. Success means the business actively uses the developed solution and the project delivers real, measurable value.
This context can sometimes be challenging or even call into question the project’s justification. That’s okay. As humans, we tend to become attached to our ideas. However, the sooner you identify these requirements, the sooner you’ll understand your potential return on investment. It’s better to end a project before development than to spend money developing something that will never be used.
To properly identify special requirements and limitations, consider these key questions:
- What data do I have that can help solve the defined problem?
- What is the volume of my data and in what formats is it stored (databases, Excel files, images, PDFs, scanned documents, etc.)?
- What security requirements must be considered for the target solution (legal requirements, country-specific regulations, industry standards, organizational policies)? Document all data regulation requirements.
- Who are the target users? Are they technically proficient or non-technical?
These questions serve as a starting point – adapt them to your specific needs. In my experience, people often underestimate both the importance of data and the legal considerations in AI-based solutions.
How to approach this? Focus on understanding the current process you want to improve or automate, including its requirements and limitations. When you can describe that process in detail, the full context usually becomes clear.
3. Don’t Follow the Trend
This may seem controversial, but many people pursue certain technologies simply because they’re trendy, because they’re widely discussed, or because competitors claim to be investing in them. Following trends is fine as long as you don’t try to force-fit problems into trendy solutions. When you redesign problems to fit technology, you risk losing sight of your original business case and spending resources on non-issues.
Currently, there’s significant hype around Generative AI. While people want to apply it everywhere, GenAI’s strengths lie in generating text, images, videos, and other content using models pre-trained on large Internet datasets. The quality of results depends heavily on the training data, and these models can sometimes hallucinate or produce poor outputs. It’s crucial to understand that GenAI (especially publicly available models) isn’t a universal solution. These models are pre-trained on specific data to perform specific tasks and produce specific results.
Experience shows that most implementations require adaptation to domain-specific needs and extensive experimentation to find optimal solutions. This process demands time and dedicated expertise from qualified engineers – without these investments, you cannot expect high-quality results that fit every use case.
That’s why defining your business case correctly, along with all requirements and constraints, is essential. This approach increases the likelihood of choosing the right technology for your problem. Sometimes the solution won’t be the trendy choice. Trends are unpredictable – while some evolve into stable market solutions, others prove temporary and offer little long-term value. Sometimes older, more stable and reliable technology is the better choice. The right decision depends on your organization’s specific needs.
4. Prepare Your Organization for Transformation
Sometimes implementing new technology in an organization isn’t straightforward, even with clearly defined business cases, requirements, constraints, and the right technology selection. Why? Because we often forget that people are at the heart of every change. Some will be end users of the technology, some may face job displacement after implementation, and others may resist because it requires learning new skills outside their comfort zone. These challenges vary, and their magnitude typically increases with organizational size.
A study by PwC shows that “more than half of workers feel there’s too much change at work happening at once, and 44% don’t understand why things need to change at all.” Implementing changes is typically easier in small organizations than in large, global ones. The more people affected by a change, the more challenging implementation becomes. However, the fundamental principle remains: clear and transparent communication is key to success.
What steps can you take?
- Identify the individuals or groups who will be affected by the change
- Determine how the change will impact them
- Prepare them for the change
The third step is often the most challenging, as it involves shifting attitudes and perceptions. Rather than “psychological tricks,” consider proven change management strategies.
5. Take Baby Steps
Sometimes the problem you want to solve is so large that the required time and money investments seem overwhelming. Questions arise like “Why does it cost so much?” and “Why will it take so long?” You might delay starting the project and eventually abandon it because you’re uncomfortable with the budget, timeline, and scope. But there’s a better approach.
As PwC’s study notes, “To get started, we need to look for ‘low-hanging fruit’ opportunities to apply AI, collect results, and use this knowledge from experiments to create the foundation for a comprehensive business transformation initiative.”
Start with something simple – it may be limited, imperfect, help only one person, or still require human involvement. That’s fine. Break your problem into smaller sub-problems. This approach allows you to quickly test hypotheses and make informed decisions about further investment. A smaller project means a more focused scope, manageable budget spread over time, and faster solution delivery. All of these factors increase your chances of success.
Conclusion
While we focused on Generative AI trends in this article, these tips apply to any AI implementation project. The key thing to remember is that your starting point should always be the problem or challenge. If you build a solid foundation, the right solution will always emerge. AI will revolutionize our world and workplaces in the coming years. To get the most out of it, think about the low-hanging fruit it can bring to your business and use it wisely. Take baby steps, experiment, learn, and see how you grow. This is the only way to avoid falling into the bandwagon effect.
Want to learn more about implementing AI in your organization? Download our comprehensive ebook “How to Implement AI in Your Company” and get detailed guidance on strategy development, team preparation, and successful implementation techniques.
Artykuł 5 Essential Tips to Avoid Generative AI Implementation Failure in 2025 pochodzi z serwisu DLabs.AI.
