(As they say, great minds think alike. I’d been kicking this draft around when Noelle Saldana reached out for feedback on a conference talk she was developing. Turns out we were thinking about the same topic, so we brainstormed together and swapped ideas. Huge thanks to Noelle for all of her help.)
I work as a consultant in the AI space. My job is to help organizations succeed with this powerful, transformative technology. I have studied patterns of how companies struggle with – or even, fail in – their data efforts, as doing so helps me identify warning signs and steer my clients clear of trouble.
One pattern I think about a lot involves an all-too-common scenario: after the initial excitement and fanfare of “becoming data-driven,” the company quickly hits a plateau on basic reports and analyses. Or they commission several projects that don’t achieve liftoff. Or, if they’re lucky, they manage to launch a few projects … but then miss several other opportunities to make the most of what AI can offer.
In short: they never reach their full AI potential.
Why does this innovation stagnation happen? What’s the common thread here? In short:
These companies treat AI as an add-on.
That rarely works out. If you want to position yourself for a win with AI, you need to be more strategic about how you employ it.
These companies never come out and say that they’re just going to tack on a side of AI. Definitely not. But they follow a familiar path:
- The leadership team decides they want to get into AI.
- They hire a data scientist or chief data officer (CDO).
- They sit back and wait (impatiently) for the magic to happen.
Company leaders often get to Step 3 with the best of intentions. Some have explained to me that they want to hire that lead data scientist or CDO and then get out of their way. They want this first hire to “own data.”
By giving them free reign, they say, they are showing respect for that person’s experience and expertise.
And sometimes they really mean it. Other times, it’s a way for the rest of the executive team to wash their hands of the whole affair if things go sour. As the saying goes: “Success has many parents; but failure, only one.”
Whatever the reason, it creates a vacuum. That’s a problem. AI by itself is just a tool; it doesn’t come with a ready-built purpose in an organization. It takes a lot of context to identify the places where it can have the most impact (develop a strategy). It then takes technical skills to build it out (execute on that strategy). And it takes a team to make all of this happen. You can’t lump all of your hopes onto a single person.
To draw an analogy, consider a home remodel.
Let’s say that several of your neighbors have added spare rooms to their homes. They seem pleased with the results and you want to follow their lead. What do you do?
You could call a builder to create an addition. If you’re lucky, this won’t violate any local building codes or cause any stress on the frame of your home. It may even match in appearance so it looks like it was part of the original structure. Maybe.
But what does that spare room actually do for you? You have more space, sure … but is that what you wanted? Is that the best you could do?
Try rerunning this scenario. But this time you’ve retained a home renovation specialist. That person learns more about your lifestyle, what you like about your home, and your future plans for the property. Together, you confirm that you need more space … but it goes beyond the addition. You need wider hallways, more room for entertaining, and a suitable kitchen for preparing large meals. You also wouldn’t mind a space off to the side to rent out. Now the plan is to knock out this wall, slide the kitchen appliances over a few feet, add a new door here…
When you’re done, you don’t just have an addition; you have a radically different home. One that works better for you day-to-day and also fetches a much stronger price when you put it on the market next year.
Does this approach take more planning and effort than just tacking on an addition? Yes. Does that planning and effort lead to an outsized return on your investment? Absolutely.
Bringing AI into your company works the same way. Sure, you could spin up a new analysis here, an ML model there, and putter along. But for maximum value, you need the bigger picture. You need to work with someone who understands AI, who can map the capabilities of this technology to your company’s goals and challenges, and who will impart that knowledge to your team.
(I emphasize that you’d work with this person or company. You definitely do not want to retain the services of a seasoned data professional – such as yours truly – and say “just tell us what to do with our data.” That doesn’t work so well.)
While AI doesn’t always live up to the hype, it can deliver results … if you approach it with the right mindset.
You’ll position yourself for a win if you treat AI as a strategic enabler. It’s a lens through which to see your company: What new capabilities does it add? Where does it drive revenue, serve as a force multiplier of productivity, or mitigate risk? To get there, you must take an active role in building out those AI capabilities.
At the start of this article, I mentioned a three-step path to mediocrity in AI. We can now rewrite that as your company’s plan to achieve its full AI potential:
- You (the leadership team) decide that you want to get into AI.
- You develop an understanding of data (“data literacy”).
- You hire a data scientist or CDO, or retain an outside consultant, to help you draft a plan for how you’ll use AI. (You bring the company knowledge; they bring the AI know-how.)
- You stay in contact with the data scientist or CDO as they execute on that plan, build out a team, and weave AI into the organization and its products.
Here you can see why the original step 3 – asking that new CDO or consultant to “own” data in the organization while you wait for magic to happen – is how you treat AI as a simple add-on to the business.
Yes, you’ll want to defer to their experience. But the continuous involvement of your key executives, product owners, and other stakeholders is key to strategically renovating the company around AI’s capabilities. Those people must continue to remain involved, to serve as internal champions of the work and to help prioritize it among other initiatives. All of this will ensure that the use of AI reflects the company’s business model and vision.