If You Don’t Think Your Product Needs An AI Strategy, You’re Wrong

Two women speaking to people in a foreign country with a robot helping them

Meeting new people and making friends in a foreign country, with help from Generative AI. Image via Midjourney.

Navigating the promises and perils of AI for your product, and how strategy design can help

“If You Don’t Think Your Product Needs An AI Strategy, You’re Wrong” originally appeared on Product Coalition.

Are you feeling pressure from your leadership to get Generative AI into your product as soon as possible?

Are you tempted to hastily shoe-horn some AI integration “hack”?

Does everyone else seem to have it all figured out, and are laughing behind your back at your lack of “AI Strategy?”

How might we understand ways to put Generative AI tech in context to provide genuine value to our users, and avoid getting disrupted in our industries by faster-moving competitors?

In this piece, we’ll unpack a better way for designing strategy to integrate Generative AI into your product.

We’ll also review one important reason why Product Managers should think twice before ever letting Generative AI design their strategies for them.

Working backwards

Understand that strategy will always be your best opportunity to go upstream and put any new tech in context for your company and your users.

But any winning strategy is always focused on solving a customer-centric problem.

It’s important to take a step back and understand that Generative AI, like any other tech, only has value to the extent it helps solve one of your target customer’s problems, in a viable way for your business.

You’d be surprised how often even established companies miss one or both of those crucial points as they get swept up in the fever pitch of Generative AI hype and FOMO.

Spoiler alert: You’re not the customer

When designing strategy with Generative AI in mind, the assumption within many organizations is to only focus on the potential productivity boosts the tech can offer, as they look to cut costs by slashing employee payrolls.

Unfortunately, one all-too common approach many companies take with Generative AI is to use it as a way to further distance their customers from speaking to their employees.

This is your chatbot speaking

During my time working at a global tech consultancy almost a decade ago, we productized and sold a full-service AI and Machine Learning-powered chatbot solution. Some clients bought and implemented them, and they ultimately had varying success depending on how well they trained and kept the models updated.

With the renewed interest in Generative AI, many consultancies have once again dusted off these turnkey chatbot solutions, prominently adding “AI” to their marketing materials.

But Generative AI-powered or not, it’s still just another chatbot.

Poorly-integrated AI can cost you more than you realize

Companies that fire their Customer Success people with the mistaken belief chatbots will do a better job for cheaper are in for a rude awakening.

Because as soon as a client really needs to talk to a human being, forcing them to speak to a chatbot instead inevitably ends up doing far more long-term harm to your brand than any short-term benefit to your bottom line.

Avoiding these traps can only come from picking the right strategic approach to the new technology.

Matching the right strategy to the technology for your users and your organization

While there are several frameworks for integrating new technologies into your strategy, there’s one that offers significant promise and a proven track record.

In his Medium piece, “Information Technology & Strategy,” Roger L. Martin introduces the concept of technologies that “flatten” or make entire industry sectors obsolete, taking with them all related and supported businesses and industries.

3 ways to integrate with “flattening” technologies

Martin shares three key ways companies can think about integrating flattening technologies into their strategies:

  1. You can invent and build out the flattening technology yourself
  2. You can build a superior early technical advantage in the new technology
  3. Or use Customer Insights to take that technology and integrate it in innovative ways in the service of creating better customer experiences

#1 is what Sam Altman and OpenAI did with ChatGPT, and not likely to be replicated anytime soon.

#2 is what Larry Page and Sergey Brin did with search algorithms and the early Web in 1998. Also not likely to be replicated anytime soon.

#3 gives us the highest likelihood of designing the right strategy to integrate with the promise of Generative AI. Easiest to replicate, as long as you heed Roger Martin’s guidance:

“For this strategy to be fruitful, a company needs to understand its customers better than its competitorsserve them with affection rather than abuse, and leverage the ubiquity of the flattening technology to set a superior standard for serving the needs of their customers.”

Roger L. Martin, “Information Technology & Strategy

The dull moments of life

My father-in-law hates to drive to LAX airport with a white-hot burning intensity, which is a miserable experience any time of day.

On a recent trip out West, we navigated out to LAX using Waze, which takes advantage of AI and Machine Learning to make navigating any drive as simple, easy, and low-stress as possible.

My father-in-law was completely blown away that a drive he’d always avoided at all costs could be as painless and stress-free as it turned out to be, thanks to Waze.

Delivering on this promise with every drive. Via Waze.com.

The power of Generative AI to ease the pain of drudgery

And that’s exactly the promise Generative AI holds — it can offer ways to take the drudgery out of many everyday experiences, enabling and empowering people to be better, faster, and more effective.

For example, take any situation where people need to sort through massive amounts of data to identify important insights to take action on. Because people tend to get tired and occasionally make mistakes, in areas like scanning for banking fraud patterns, even small human errors can end up costing banks untold billions of dollars. But Generative AI never gets tired, and can comb through mountains of data and pinpoint anomalies rapidly without error.

But what about Generative AI’s massive potential to create totally new creative content, from articles, screenplays, and marketing campaigns, to music and images (like the one at the top of this post)?

From “Generative AI: A Creative New World,” via Sequoia Capital.

The biggest promise and threat of Generative AI — new content creation

And this really is the biggest shift Generative AI represents — for the first time, we have technology that can be prompted to create something that’s never existed before.

But keep in mind that Generative AI will never make people obsolete. They’ll only be made obsolete by other people who know how to skillfully prompt great Generative AI-powered products to take advantage of its powerful “multiplier” effects.

For better or worse, Generative AI’s incredible potential for content creation is so feared that it’s at the center of the currently ongoing writers & actors dispute that’s now turned into the longest labor dispute of its kind.

While much of that potential is as of yet untapped, at its worst, Generative AI remains largely unreliable, still prone to “hallucinations.”

And when not skillfully prompted, it typically writes weak, trite, and flat-out factually incorrect garbage which clearly wasn’t written by anyone who cares, nor fit for human consumption.

When the content is only part of the value

“What you get by achieving your goals is not as important as what you become by achieving your goals.”

Zig Ziglar

Yet many today sell books and courses touting the potential of ChatGPT to rapidly churn out tons of SEO-friendly content even as it threatens the livelihoods of journalists, bloggers, marketers and social media strategists everywhere.

But this misses the point that in much of writing, the content you end up with only represents a small part of the value of the creative process.

The benefits of digging deep

The true benefit of much writing is the amount and depth of research and Type 2 thinking you have to do to create it, and how it forces you to stretch, learn, and grow in the process.

This is absolutely essential not only for students who need to develop judgment and critical thinking, but Product Managers writing Product Requirements Documents (PRDs), User Stories, Jobs to Be Done, or Amazon-style 6-pager narrative memos.

And the same goes not just for individual creators, but doubly for crucial team-based efforts like strategy design or goal-setting.

The benefits lie in the process

I would argue you would never want to input your “requirements” and have some Generative AI-powered platform tell you what your strategy should be, or what Objective and Key Results (OKRs) you should be targeting.

Because the biggest advantage comes in building a shared experience, what you and your team are forced to learn, and how you’re all collectively encouraged to stretch, expand, change and grow through the shared exercise.

Few group activities can equal the collaborative struggling through the iterative Diverging and Converging waves of creating, throwing away, changing, and ultimately deciding on the right strategy to move forward with as a group.

It’s the process itself that not only builds and sustains teams to better collaborate and make better choices together, but gives them a head start in putting that strategy into action.

Strategy will always be your best way through

Your best bet to integrate Generative AI into your product?

Taking our cue from Roger L. Martin’s point #3 above

  1. Start by gathering deep customer insights learned through continuous generative interviewing.
  2. Pull together a cross-functional group, including perspectives from:
  • Business
  • Product
  • UX
  • Technology

As part of your technology people, make sure you include Subject Matter Experts (“SMEs”) with direct, hands-on experience integrating with Generative AI and Machine Learning and related technologies

Strategy design as a social activity

Use this group to facilitate stepping through the 7-step Strategic Choice Structuring Process to design your strategy as a team.

The Strategy Process Map. Author image from Roger L. Martin’s copyrighted work.

You’ll spend most of your time at Steph #3 above, generating sets of Strategic Choice Possibilities that offer promising and innovative ways to create great customer experiences where Generative AI will play an important role as either a “How to Win,” a “Must-Have Capability,” or an “Enabling Management System.”

For each set of strategic choice possibilities, the key will come in asking

“What would have to be true?” (“WWHTBT”)

What would have to be true about…

  • Our customers? Will they want to interact with Generative AI in these ways?
  • Our company?
  • Do we have or can we build the “Must-Have Capabilities” to be able to build and sustain these innovative integrations at scale?
  • Do we have the “Enabling Management Systems” to measure and continuously improve those capabiilties?
  • Our competitors? How will they react? Can they easily copy how we’ve integrated Generative AI to neutralize our differentiating competitive advantage?

Once you’ve raised these questions at Step 4, you will design tests at Step 6 around your biggest potential barriers before collaboratively choosing a strategic choice set as a way forward.


Any strategy approach to integrating Generative AI into your product will always come back to one thing: how well you solve your target customer’s most urgent needs in unexpected and delightful ways that are margin-enhancing in ways that your competition either can’t or won’t be able to easily copy.

Your best way to do that lies in understanding your customers better than the competition, and collaboratively and cross-functionally designing your strategy to integrate the technology in innovative ways to make the most of those opportunities.

Join my newsletter list for the Upstream Full-Stack Journal, connecting the dots on the full Value Delivery stack, from Strategy to OKRs, Product Management, through Agile Systems of Delivery


Roger L. Martin, “Information Technology & Strategy”


Sequoia Capital, “Generative AI: A Creative New World”


Farnam Street, “Daniel Kahneman Explains The Machinery of Thought”



%d bloggers like this: