Continuous Transformation Blog

AI Adoption: Constellations, Not North Stars

Written by Neil Sheppard | May 16, 2024

Generative AI adoption is not a straight path and putting all your eggs in one basket may not be the right choice. Find out how you can leverage a variety of AI strategies to mitigate your risk.

Industry commentators insist that companies need a clear strategy for adopting and deploying artificial intelligence (AI) in the enterprise. While clarity is indeed critical, it's important to remember that clarity doesn't mean picking a single approach and sticking with it no matter what..

The simple truth is, there is no one, best way to utilize this brand-new technology. For this reason alone, it's best to keep your options open rather than rolling the dice and hoping your AI strategy turns out for the best.

This doesn't mean you should hold back on seeing what you can do with AI technology, however. Avoiding AI is, itself, a single-minded strategy that might not pay off.

The way forward with AI adoption is to explore several different AI strategies, without fully committing or risking vendor lock-in. As time goes by and the future of business AI becomes more clear, you'll be in the best position to move forward.

Keeping your AI options open starts with understanding exactly where you are using AI solutions in your company today. To do that, you need an application portfolio management solution, like SAP LeanIX, with functionality dedicated to supporting AI adoption and governance.

 

Lost In The AI Fog

Forrester believes that generative artificial intelligence (AI) will be as transformative as the internet, the smartphone, and social media. Just as none of us could have predicted the rise of Netflix, Uber, DoorDash, or TikTok, it's impossible to predict exactly where AI will take us in the future.

This unpredictability puts tremendous pressure on IT leaders. Laggards in AI adoption risk being left far behind by the competition. Early adopters, on the other hand, could be big winners. Of course, they could also end up locked-in with the wrong vendor, or worse, find they've built their AI strategy on fly-by-night vaporware.

It's like being lost in a fog and knowing a step in any direction could lead you off the edge of a cliff. Since companies feel compelled to move despite the uncertainty, IT leaders and others are casting about for clear direction.

We find some companies, such as Dropbox and Deloitte, seeking to create clarity simply by being decisive. They advocate choosing a 'north-star metric' or key goal that will guide your AI strategy wherever the market may turn.

Sticking with our 'lost in the fog' metaphor, is essentially like picking a direction, walking that way confidently, and hoping it's the right one. This is a great strategy for overcoming 'analysis paralysis', but it also places the future of your business in the hands of lady luck.

 

Is An AI "North Star" Really The Right Strategy?

Being somewhat arbitrarily decisive is a common strategy when it comes to complex situations. Not taking action will often lead to a worse outcome than making a mistake, so it's best to just do something, rather than run out of time trying to work out the right thing to do.

If one accepts the premise that AI will ultimately be game-changing, and that those who successfully harness AI will quickly best their competitors, 'doing nothing' falls away as a reasonable option. In that light, choosing a single north-star metric – which doesn't necessarily mean locking yourself into one strategy – has a definite appeal.

But, what if you choose the wrong metric? If the metric you select is customer-facing, say, NPS, but it turns out that AI is better suited for improving internal process efficiency or something like accounts payable, then you were barking up the wrong tree.  

Since we really don't know what the best-practice use case(s) for generative AI will turn out to be, you're better off experimenting with the technology in several use cases and seeing what works. This approach allows you to uncover the paths to value for your business while helping you avoid getting trapped in a dead-end strategy.

 

AI Constellations, Not North Stars

Rather than focusing on a single, north-star metric, start tracking the potential use cases for generative artificial intelligence (AI) in your organization. Derive a set of metrics for each use case, track, and analyze.

Of course, while you are finding the best approach for your business, you need to ensure that your employees aren't using AI in ways that will cause security or reputational risks, such as when Samsung caught employees uploading proprietary code into ChatGPT. That being said, it is worth giving employees the chance to test out AI solutions in safe sandbox environments to discover their value.

When the results of your experiments look promising, increase investment in that area. When a certain use case has been explored and isn't offering a return on investment, then end the experiment and explore other options.

This is, of course, more effort than simply committing to an AI strategy that feels right, and you will need to carefully limit your investment so the costs don't add up. However, by diversifying your AI adoption strategy, you gain empirical evidence of what AI tools have value to your organization for particular use cases.

By remaining agnostic about AI, developing business use cases, and taking tentative steps forward with tools that offer real value, rather than jumping in with both feet, you can unlock all the value that AI promises without exposing yourself to potentially devastating risks. This is the only safe strategy for leveraging AI, especially since refusing to implement AI tools could have negative consequences in the long term.

 

Leveraging SAP LeanIX For AI Adoption

SAP LeanIX Application Portfolio Management enables you to track AI, its usage, and its risks across your IT landscape. It also allows you to share this information with all your stakeholders through data-rich dashboards.

Using SAP LeanIX Architecture and Road Map Planning, you can then experiment with this data in a sandbox environment to see what potential changes you could implement across your architecture going forward. Once you have a vision for the future, you can then map a path from where you are to where you want to be.

To find out more about how LeanIX products can support your AI adoption journey, book a demo: