What NOT to Do When Identifying AI Use Cases for Your Business

What NOT to Do When Identifying AI Use Cases for Your Business

The time has come. You’re ready to get serious about AI adoption in your organisation. The first order of business: identifying use cases. Seems simple, right?


Identifying an applicable use case that aligns with business goals is vital for your organisation’s AI adoption journey. But with the rapid pace of technological advancement and business leaders’ lack of AI understanding, it can be difficult to know where to start.

When it comes to pinpointing viable use cases for AI, many business leaders fall into the same traps, ultimately leading to a waste of resources, time, and money.

To help you avoid these pitfalls, we’ve made a list of what not to do when identifying AI use cases for your business:


  • Forget business and technology alignment

Before any AI adoption takes place, you need to make sure any AI project you’re about to spearhead aligns with your overall business strategy and objectives. Any digital transformation initiative should have the company’s wider, overarching goals at the forefront and be built upon from there.

Aligning business and technology means your AI project should be strategic. This involves setting objectives, identifying your KPIs and metrics, and tracking ROI. The whole point of AI adoption is to bring value to your business, not just implement tech for the sake of it.

Something also worth mentioning here is the danger of falling into a bandwagon mindset.

Many business leaders have heard of technological solutions that others in their industry are adopting and want to implement these same solutions for their organisation. However, they are usually unaware of the challenges and limitations of these solutions when it relates directly to their business (eg. data requirements, existing systems, and processes, etc.)

Each business is different. Make sure your solutions meet your business’s unique needs and fit your business’s unique requirements.


  • Overlook legacy systems and processes

Does your current infrastructure and environment support AI integration?

What existing processes and workflows are in place and how would they be affected by AI/ML adoption?

These are a few of the questions you should ask yourself to assess the capabilities of your current legacy systems for AI integration.

When referring to “legacy systems”, we refer to any technology, a computer system, business processes, or an application that is outdated compared to the most modern standards and expectations or current IT contexts. Legacy systems are referred to as such not because of their age, however, but for their inability to meet an organisation’s current needs.

At the core of modern-day AI is machine learning, which relies on models that are distinct from simplistic process-based task models. Most organisations’ legacy systems are built upon these process-based models and cannot be easily integrated with ML systems.

When identifying business cases for artificial intelligence, take into account that new applications, processes, and workflows will be needed to optimise the results (and the budget to do so). Force-fitting an AI solution to a company’s legacy system and employees who are not primed for adapting is a surefire way to fail.


  • Overlook data requirements

AI depends heavily on data, and an AI solution is only as good as the data that is being fed into it.

A common pitfall business leaders make when brainstorming AI pilot projects is forgetting about data availability. Do you have the data needed to build the AI application you want? If not, is there a viable way to generate this data, be it real or synthetic?

When considering the data needed for a viable AI solution, remember that quality is just as important as quantity.

As a general rule, the more data available, the better. However, if the data you are analysing will not bring business value, it’s as good as useless. It’s far better to have a smaller quantity of data that is directly relevant to generating business value than truckloads of useless information. As we explained in a previous blog, less is more.

If your organisation lacks the necessary data for an AI business case, or cannot release specific data points due to regulation and privacy concerns, synthetic data sets can also be generated by your in-house data scientists or an outsourced AI vendor.

Moral of the story: Business cases for AI should be ideated while simultaneously checking access to relevant data sources. If you aren’t doing this, it’s likely you’ll have to return to the ideation phase having wasted time, resources, and momentum.


  • Run before you can walk

This point is especially relevant to first-time AI adopters.

It’s easy to become overly optimistic about the impact company-wide AI can have on any organisation. However, when dipping your toes into the world of enterprise AI, don’t go too big, too fast. The first pilot should be small(ish), and needs to bring results fast.


Firstly, it ensures that you mitigate the risks of costly mistakes. A quick return on investment is especially important for smaller and medium-sized businesses that don’t have large budgets to reach from.

Secondly, the first AI project is a testing ground, allowing stakeholders to see how this technology works inside the organisation.

Thirdly, if you have a successful model that brings tangible results and business value in a short period of time (think anywhere under a year), you have leverage. Buy-in for future applications of AI will be much easier when it comes time to persuade someone to invest in larger, more time-consuming projects.

Now, this isn’t to say that you should toss the idea of a scalable model out the window. Identifying use cases that can be scaled across your business is a smart move and should be considered. Use the first project to test the model, and if successful, scale it throughout relevant processes after.




Introducing AI into your business is a worthwhile investment, but only if done properly. Failing to keep your organisation’s current processes, workflows, data sets, and overall strategy in mind while ideating AI use cases will ensure failure.

Are you considering AI adoption in your enterprise but having difficulty in defining viable use cases? Nexus’ AI Factory is a pilot production centre for companies where they can discover, evaluate and scale AI opportunities within their organisation. Get in touch with a consultant today to find out how the AI Factory can help you accelerate AI adoption in your business.

More articles