Innovation is a crucial element of many modern businesses, particularly SMEs and solopreneurs looking to differentiate against larger competitors. Artificial intelligence can provide business innovation, but can also drain budgets for little useful return. Here are some tips to help business leaders use AI effectively.

Use AI for the Right Reasons

Separate the purpose of your business project from the reason you’re using AI, and make sure AI is not only necessary to achieve it, but the best means to doing so. Marketing departments can be keen to claim a product is “powered by AI”, but can find it harder to explain how that translates into benefits for a customer. Business leaders can be guilty of the same thing, wanting to use AI in projects for the sake of using AI.

Part of the difficulty in deciding whether AI is the right technology to solve a business problem is that it’s not necessarily clear what AI actually means, and boundaries with other technologies can be blurry. AI covers a diverse range of applications, several of which don’t have much in common. AI is often used to analyze data and forecast likely consequences of decisions, although this is also the territory of “Big Data”. Another common use is the automation of business activities that would previously have needed human input, for example, posting transactions to ledgers. Computer vision is a familiar type of AI, for example, recognizing objects or faces in-camera images. A version of this is document management, in which the contents of printed and hand-written documents are scanned and converted into computer content. Closely related to this is natural language processing, whereby computers extract the meaning from the text as well as audio speech. Robotics is the final type of AI in common use, from robot vacuum cleaners to self-driven cars.

In practice what you call the technology doesn’t matter as long as you have clarity on what you want to use it to do. Which brings us to the next tip.

Use AI for the Right Problem

There is almost no limit to the ways AI could be used in large businesses, and even SMEs have many potential opportunities to use AI. However, the reality is that most of the time it’s used to address a handful of types of problem.

One of the most common is improving our understanding of customers, both individually and collectively. AI can help generate insights about products customers prefer and ways they like to choose and buy them. It can also flag early signs of customer dissatisfaction and potential churn.

A less specific use of AI is making better decisions that rely on data. The type of decision can vary tremendously, but can cover anything from shortlisting CVs of prospective employees to setting the optimal discount to offer a customer. The key is decisions that can be made or improved by sifting through lots of data.

AI can be important in product development, potentially adding competitive advantage to features of a product or service. Many startups, with the advantage of agility and speed, have come to market on the basis of using AI to offer something larger, more established competitors can’t provide.

One of the biggest uses of AI in larger firms is intelligent automation, sometimes known as RPA (Robotic Process Automation), and allows AI to introduce step changes in efficiency, speed, and scale of tasks.

There are many other possible uses of AI in business, and an internet search for “AI Use Cases” will give you a sense of these. But most examples cater to large corporates with deep pockets, and smaller businesses need to approach AI differently from larger ones. An important starting point is deciding whether to build or buy, which is what the final tip is about.

Don’t Build AI Unless Really Necessary (Buy or Rent Instead)

Much advice for business leaders about AI focuses on overseeing the creation of effective AI business solutions. However, as the AI market has evolved, the option of buying pre-built AI solutions or configuring part-built AI applications has become increasingly common.

In fact, the introduction of AI features into commonly used software means many businesses have already started using AI without realizing it. For example, many CRM applications routinely use AI to help users target customers, design campaigns, and even draft emails. This means that the distinction between an AI business project and a regular technology business project is diminishing. As a result, once you have clarity on what your business problem is and how you want to use AI to solve it, you may well find the features you need have already been incorporated into mainstream software used in the same business area.

If that’s not the case, it’s also possible that a separate AI solution has already been built, and your AI business project is then about evaluating pre-existing solutions and adopting the one that most closely suits your needs. Most of these are likely to be available through cloud subscriptions, so it may even be possible to try different options and select based on actual use.

Was Rahman is an expert in the ethics of artificial intelligence, the CEO of AI Prescience, and the author of AI and Machine Learning. See more at www.wasrahman.com