The right way to make AI part of your tech strategy

Estimated reading time: 8 minutes

These days, AI is everywhere. Depending on who you ask, it’s either poised to automate all of software engineering by next quarter, or it’s just the latest overhyped buzzword.

But for anyone who’s used AI tools beyond a quick demo, the truth lies somewhere in between.

Engineering leaders now face mounting pressure to reshape their technology roadmaps in response to AI. Not just because of a desire to capitalize on the latest trend but because business stakeholders see enormous potential: more productive teams, new markets, enhanced products, predictive insights, and hyper-personalization at scale.

Strategic use of technology has always been core to business success. And defining a clear technology strategy ensures engineering investments align with those business objectives. Without one, companies risk falling behind or chasing the wrong trends.

As AI dominates many boardroom discussions, knowing how to align innovation with your business model isn’t just advantageous. It’s existential. 

With 25 years of experience using AI and exposure to leading a startup building AI-powered tools, I’ve found that navigating between strategic innovation and executive urgency can be less complicated than it seems.

When technology becomes the strategy (instead of supporting it)

Throughout my career, I’ve seen many businesses fall into the same trap: treating technology itself as the strategy, rather than as a tool to enable it.

Think back to the early 2000s, when “digital transformation” became the buzzword of the decade. It promised a sweeping revolution in how companies operated: more efficiency, better customer experiences, and data-driven decision-making. Executives poured billions into new tools, hired chief digital officers, and launched ambitious initiatives. And when implementation outpaced understanding, teams were left grappling with technical debt, internal resistance, and solutions in search of a problem. Sound familiar?

We’ve seen the same pattern repeat with each wave of innovation: cloud computing, microservices, Kubernetes, blockchain, extended reality, and, now, AI. 

Take cloud computing: despite widespread adoption, many companies are now facing rising costs, vendor lock-in, and unexpected complexity. The troubles have been so extensive that 42% of U.S. organizations are moving at least half of their cloud workloads back on-prem, a recent Citrix study finds. 

These patterns raise a hard truth: simply adopting cutting-edge technology doesn’t guarantee results. In fact, companies that center their strategy around the tech itself (rather than how it serves their specific business goals) often end up worse off.

So the question is: is AI just another overhyped trend, or will it bring concrete innovations?

While AI still holds massive potential, getting from potential to payoff isn’t automatic. We haven’t yet seen broad productivity gains outside of a few niche domains. An IBM survey found that three in four AI initiatives fail to deliver their promised ROI. A recent study from the National Bureau of Economic Research tracked 25,000 workers in AI-exposed industries. The result? Minimal gains in productivity, and almost no impact on wages or hours worked.

So yes: AI should be part of your tech strategy. But not everything needs an AI solution.

Like any tool, it’s only as good as the context in which it’s used. Technology, no matter how advanced, is not a strategy on its own. Real success still depends on people, process, and a clear business purpose.

The “AI-first” backlash and the quiet self-correction

Several companies that raced to declare themselves “AI-first” are now walking those announcements back. From Duolingo’s CEO, Luis von Ahn, making headlines by proclaiming the company would become “AI-first,” which is messaging the company had to later soften, to Shopify’s internal memo that suggested AI-driven productivity would reduce headcount, there has been no shortage of public examples. 

And there’s more. In April of this year, Klarna followed a similar arc. The company initially celebrated the cost-efficiency of its AI chatbot, claiming that it hadn’t hired humans in a year. But the tune changed quickly when the chatbot’s “lower quality” led Klarna to start hiring again.

Each of these stories reveals the same underlying truth: bold claims about being “AI-first” might grab headlines, but they often underestimate both the complexity of real-world implementation and the human factors that drive product quality, trust, and long-term success.

The bias toward action

AI seems to promise everything: competitive edge, operational efficiency, new business models. And when revolutionary tools hit the market, it’s natural for organizations to feel they must act fast or be left behind.

But this urgency often stems from a cognitive trap: our bias toward action.

When faced with uncertainty, we’re wired to “do something.” We equate movement with progress. After all, we’ve been conditioned to associate effort with results. So the instinct is to build, deploy, announce. But movement is not the same as velocity.

• Speed is how fast you’re moving.
• Velocity is how fast you’re moving toward a goal.

You can run fast in circles and still go nowhere.

This is especially true with AI adoption. The pressure to act – often driven by stakeholders, competitors, or media hype – can lead to hasty rollouts with unclear objectives, missed risks, and wasted resources. Action without strategy becomes luck at best and costly at worst.

Instead, what’s needed is strategic patience. In fast-moving domains, it’s tempting to believe the only options are to act now or to fall behind. But there’s always a third option: pause, assess, and get directionally aligned.

Gather information. Explore use cases grounded in your business reality. Align on goals. Then act intentionally.

As Henry D. Thoreau wisely wrote, “It is not enough to be industrious; so are the ants. What are you industrious about?”

Where AI investments make sense

The key to integrating AI into your technology strategy is simple but often overlooked: use it to solve a real, proven problem.

For engineering leaders, the best opportunities for AI are in removing bottlenecks, reducing toil, and enhancing what your teams and users already do.

Here are some places to start before you invest: 

1. Listen to your builders.
   
   Your engineers (those experimenting with generative AI, integrating LLMs, and debugging with AI copilots) understand both the possibilities and the tradeoffs. They can offer critical insights into whether AI adds value in the context of your product and architecture, or if it introduces new overhead in process, tooling, or trust.
   
2. Anchor in user pain.
   
   AI features that succeed typically start with deep user empathy. Ask: what problem are we solving? For whom? Would solving it with AI materially improve their workflow, outcomes, or experience?
   
3. Enhance user experience.
   
   AI should amplify what users can do, not complicate it. The best applications are invisible, elegant enhancements, not feature bloat.
   
4. Make data your advantage.
   
   AI without good data is like a jet engine strapped to a paper plane. Invest in infrastructure that lets you collect, structure, and learn from the data you already have safely and responsibly.
   

As the CTO and co-founder of a devtool startup, my guiding principle is not AI-first, it’s problem-first. During a recent webinar, I walked through how we’ve added AI to our platform in ways that support debugging and documentation – two areas developers already struggle with, and where context and automation make a meaningful difference.

Rather than asking “how can we add AI?”, we asked “where are people still losing hours of their lives doing manual, frustrating work?” And we built there. That approach helped us stay focused, avoid hype distractions, and ship features that our users actually love.

Setting expectations with business stakeholders

If you’re an engineering leader navigating when and how to incorporate AI into your technology strategy, you’re likely also managing pressure from stakeholders who want fast results and bold AI bets. One of your most critical responsibilities is helping them understand the realities behind the hype.

Here’s how to approach those conversations:

1. Highlight the risks of going “AI-first” without a strategy

AI isn’t a shortcut. It’s a tool. And misusing it can backfire. Help stakeholders understand the potential pitfalls:

• Customer experience vs. tech novelty: building for the sake of AI, instead of user needs, can hurt product usability and damage your brand.
• Operational complexity: AI rarely drops neatly into your workflows. It often creates new maintenance burdens, cost centers, and quality concerns.
• Team disruption: overpromising AI-led productivity gains can lead to premature layoffs or internal friction. Both of which harm morale and reputation.

2. Ground ideas in your specific product, market, and stage

Business leaders often come from different industries or domains. They may expect AI to work the way it did in a previous context. It’s your job to explain what’s different about your users, data, and architecture. Use benchmarks from similar companies, share lessons from failed implementations, and push for evidence-based decision-making before investing time or talent.

3. Make the case for focus

When all eyes are on the future, remind your team what innovation really looks like: clarity, not chaos. A thousand ideas won’t beat one that works.

As Steve Jobs put it: “People think focus means saying yes to the thing you’ve got to focus on. But that’s not what it means at all. It means saying no to the hundred other good ideas that there are. You have to pick carefully. I’m actually as proud of the things we haven’t done as the things I have done. Innovation is saying ‘no’ to 1,000 things.”

Sometimes, the most visionary thing you can do is double down on what’s already working and save the experiments for when the time (and problem) is right.

Final thoughts

AI is not a silver bullet. It’s a powerful tool that, when applied thoughtfully, can meaningfully accelerate product development, improve user experiences, and unlock new business value.

But adopting AI for its own sake rarely ends well. I’ve learned that the most impactful applications of AI come from solving real, validated problems, not from chasing trends. As engineering leaders, our job is to stay curious, stay grounded, and help our organizations move not just fast, but forward.