When it comes to AI, will you be predator, or prey?

If you don’t make proactive choices now, you’ll lose ground to your competitors.

“a scientific illustration of a Timber Wolf, similar to the hand drawn illustrations found in Charles Darwin's journals“ / DALL-E

You’ve been told artificial intelligence (AI) will change everything, but what does that actually mean for your business? Which AI technologies will have the biggest impact on your bottom line? And how can you avoid investing in empty promises?

“Our board wants to know our AI strategy, and I’m not sure where to start. I need some real answers.”

Here’s the truth: 

The reality is, every functional area of your business will be disrupted by AI technology over the next five years, and if you don’t make proactive choices now, you’ll lose ground to your competitors.

So, how will you avoid falling prey to a more aggressive AI adopter? In what ways are you adapting and what’s your plan to infuse this technology into your organization? How are you deciding where AI is a strategic differentiator and where it’s table stakes?

“How can we leverage AI to positively impact our P&L? What are the pain points in our business that AI could improve?”

This three-step process will help you figure out where to begin.

The most common question I’m asked by senior executives today is, “Where do we start?” As with almost any successful endeavor, the first step is to be super clear on your goals.

When it comes to AI, your goal should be to start implementing it in the places where it will deliver the highest ROI. This might seem obvious, but it can be easy to become distracted by all the hype and promises, and forget to focus on ROI.

When I’m working with a business, I like to help them build an initiative inventory with rank ordering based on a straightforward, three-step process below. (Note I said “straightforward”, not “simple”!) Here’s a good way to start:

1. First, list out all the functional areas of your business: Start by identifying all functional areas in the business and estimate their size in terms of their contribution to the P&L, typically in thousands of dollars.

2. Then map those functions to any current and potential AI initiatives you have in the works: Link both internal and vendor-led AI initiatives to their corresponding functional areas. This includes assessing current vendors' AI capabilities and their future plans. Along the way, add previously unidentified initiatives and vendors.

3. And finally, rank those initiatives based on the actual ROI they’ll deliver for you: Exponential ROI (EROI) is a R.I.C.E.-inspired metric I use to rank order the impact of an artificial intelligence initiative while accounting for exponentially growing technologies and evolving business dynamics. (See more on how to calculate this metric below.)

There's a lot more to it, of course, and each of those steps involve quite a bit of legwork, but if you follow this formula, you'll be out in front of the pack, ready to pounce on the right opportunities for your business.

The section below will show you more about what those three steps will do for you, what kind of data you should have by the end of that process, and exactly how to make it happen!

When you finish the exercise above, you should have a series of documents summarizing each initiative and all your assumptions, as well as a summary table like Figure 1 below.

Figure 1: Summary of AI Initiatives

Here are a few things to notice in the table above:

• Your final summary should list all of your AI initiatives by functional area, ranked in descending order by EROI.

• Typically companies start at the top, funding initiatives above an EROI threshold until cumulative investment matches budgeted investment.

• The four inputs to EROI shown in the table allow you to incorporate effects related to time and the team or partners ability to deliver.

Now let's look at how to calculate Exponential ROI.

The EROI calculation is simple, but the inputs are subjective and require thoughtful debate. This is a SWAG (scientific wild-ass guess) so don’t get too deep into the weeds of estimation. The four inputs for this metric are:

• Contribution: The contribution to profit and loss of the functional area (or sub-area), typically in thousands of dollars.

• Impact: The estimated percent improvement the AI initiative will have on the functional area. This is expressed as a percentage, e.g., a 10% improvement.

• Confidence: Your level of certainty in the ability of your internal team or vendor partners to deliver the estimated impact for the estimated investment. This is also expressed as a percentage, e.g., 60% confidence.

• Investment: The size of the investment required, including human capital, typically in thousands of dollars.

These inputs are then entered for each initiative into the following formula and the results sorted descending:

The “Exponential” in EROI reveals both risk and opportunity.

Artificial intelligence is advancing at an exponential rate - think of Moore’s Law on steroids - which needs to impact our input decisions. As AI’s capabilities rapidly grow, Impact and/or Confidence may be significantly greater 12 months from now. Similarly, declining compute costs and more efficient AI architectures means Investment may be significantly lower 12 months from now. The strange result is that it may make sense to defer a modest ROI project for 12 months to take advantage of the unique economics of exponential growth.

AI is a disruptive technology, so be careful not to underestimate the Impact of these initiatives. Startups and competitors are going to take them on even if you don’t and we want to avoid the Innovator’s Dilemma. While it can be hard to believe, some AI initiatives will result in unprecedented productivity gains.

“Every vendor I’m talking to is telling me how their new AI offering is going to change my business—and every one I've tried so far has not changed my business.”

A final aspect to consider when defining your Exponential ROI inputs is vendor misinformation. Frankly, marketing and sales efforts driven by breakthroughs in the underlying AI technology have outpaced product development. Every vendor is touting their AI capabilities and making grand promises. Their failure to deliver on these promises is causing many businesses to lower their Confidence for vendor-driven initiatives. Carefully verify vendor claims of both the nature of their solution (e.g. is it actually AI) and the stated benefits.

The time to start is now.

Remember, the longer you wait, the further behind you'll be. So what questions, criticisms, or suggestions do you have? Do you need help taking the first step? Please feel free to engage with me and the rest of the community in the comments below.

Postscript.