Three strategies for successfully implementing AI

One year on and we’re well and truly in the AI era. For businesses, the application of AI in a commercial setting is a major technology shift. From my perspective, just as businesses that failed to adapt to the internet got left behind, so will those that don’t adopt AI.

The problem is that with all of the hype, many business leaders have little idea where to start. In this year’s machine learning, AI and data landscape, there were at least 1,400 logos across dozens of categories, according to data compiled by a team at FirstMark, a venture capital firm. AI companies are popping up everywhere, each hoping to capitalize on growing commercial enthusiasm. So, what’s a business leader to do?

There aren’t many people who can say they’ve worked in the AI space for nearly ten years, but I can, and I’m here to be your signal through the noise with three actionable pieces of advice on getting started with AI.

Don’t let data hold you back

I’ve lost count of the number of times I’ve heard, “I can’t start yet. I don’t have my data sorted out.” I get it; data is the lifeblood of AI, but huge digital transformation projects are costly and can take years to complete. You really don’t want to wait five years to see any return on investment.

I recommend starting with a functional area that is both complex and data-rich. There will probably be some work to be done to ensure your data is AI-ready, but it won’t be anywhere near as painful as a business-wide transformation.

By using this modular approach to build out your AI capabilities over time, you can see a faster return on investment. Then, you can identify the next business area that will benefit from AI at the same time as building in-house expertise to support future AI projects.

Work with the outcome in mind

We’ve all been swept up in the hype of AI, and the temptation to jump in feet first is real. Before you take the plunge, though, you need to truly understand the problem you’re trying to solve or the opportunity you’re trying to address.

The simplest way to do this is to ask yourself what the outcome you’re trying to achieve is. For example, do you want to reduce the amount of working capital tied up in safety stock? Do you want your customers to have a more seamless experience talking to your chatbot?

Everyone wants to jump on the generative AI bandwagon. Its ability to produce novel text, imagery and audio lends itself well to content creation or image-recognition tasks, so there are real business benefits.

But there’s another type of AI that has been modestly serving business needs for years: predictive AI. Predictive AI typically uses structured data sets, like inventory logs or pricing data, to deduce the likely future or categorizes based on patterns. Its optimal use cases are demand forecasting, customer churn prediction and price elasticity, the type of areas that when optimized by AI can be game-changing to business fundamentals.

Understanding your desired outcome and the type of AI that will help you achieve it will make the landscape of possible solutions much easier to navigate.

Think longer term

Let me ask you a question: Would you cut and paste someone else’s website and use it for your business? Maybe you change the color scheme a bit, tweak a few words and add your logo? You’re likely shaking your head at me, “No, Richard, don’t be a fool. My website is the storefront of my business. It’s vital that it represents my brand and works exactly how I need it to.”

You’re right. Every business is unique, with different target customers, supply chains, people, practices and more. Each has its own set of challenges and opportunities.

Think about AI in the way you think about your website. To really make a difference, AI has to understand your business and leverage its uniqueness. To gain a true competitive advantage, you need your own AI.

That doesn’t necessarily mean you need the in-house capability to build it. The idea of developing your own AI can sound laborious and long-term, which is what makes point solutions — services, tools or products that solve specific business problems — so attractive.

Fast forward five years’ time, though, and you might find yourself picking apart dozens of point solutions that have become obsolete as new technologies have developed. Instead, look for tools you need to develop, and adapt your own AI solutions over time.

One year on from AI’s “iPhone moment,” there’s still immense pressure for business leaders to jump on the AI bandwagon. So, as you sit down to work on next year’s strategy, don’t let the thought of AI overwhelm you. Pick one data-rich area, understand the desired outcome, and find a solution that allows you to expand your use of AI over time. That way, you’ll be setting your business up for a long and happy future.

This article was originally written and published for Forbes Business Council on 12 December 2023.