Stop the AI washing and start measuring what actually matters.

The AI Pressure Cooker

“We need AI in our product roadmap yesterday.” Sound familiar? You’re sat in yet another exec meeting where someone’s waving around headlines about AI productivity gains, whilst you’re quietly wondering if anyone’s actually seen these mythical improvements in the real world.

The Reddit product management community recently erupted over whether AI is the next dot-com bubble or the next cloud revolution. The debate was fierce, but the underlying frustration was clear: product managers are drowning in AI hype whilst desperately seeking practical ways to evaluate what’s actually worth investing in.

When AI Becomes All Sizzle, No Sausage

Here’s what’s really happening behind closed doors. You’ve got executives demanding AI integration, vendors promising transformational productivity gains, and you’re left holding the bag when the quarterly numbers don’t add up.

It’s like being asked to add truffle oil to every dish on your menu because it’s trendy, without anyone bothering to check if it actually makes the food taste better – or if customers are willing to pay the premium. You end up with an expensive ingredient that looks impressive on paper but leaves everyone wondering where the promised magic went.

The numbers tell a sobering story. A recent NBER study found that many enterprise GenAI pilots showed zero or negative returns. Meanwhile, companies like OpenAI are reportedly losing $2 for every $1 they bring in revenue, suggesting those ‘all-you-can-eat’ pricing models won’t last forever.

But here’s the thing that’s keeping you up at night: you’re getting pressure to move fast on AI whilst simultaneously being held accountable for ROI. It’s a recipe for disaster if you don’t have a proper framework for making these decisions.

The AI Investment Recipe That Actually Works

Stop treating AI as a binary decision. The question isn’t “Should we do AI?” but rather “Which AI investments will measurably improve our operations?” Here’s your practical framework:

Step 1: Define Your Success Metrics First

Before you entertain any AI pitch, establish exactly what you’re trying to improve and how you’ll measure it. For retail and restaurant operations, this might be:

• Inventory turnover rates
• Customer wait times
• Staff scheduling efficiency
• Demand forecasting accuracy

If a vendor can’t tie their solution to at least one of these concrete metrics, politely show them the door.

Step 2: Distinguish Between AI Types

Not all AI is created equal. The Reddit debate revealed a crucial insight: specialised AI (predictive analytics, computer vision for quality control, demand forecasting) has proven ROI, whilst general AI (chatbots, content generation) remains largely speculative for most businesses.

Focus your investment on narrow AI that solves specific operational problems. Save the flashy GenAI experiments for when you’ve got surplus budget and time.

Step 3: Pilot with Exit Criteria

Set up every AI pilot with clear success metrics and a defined timeline. If after three months your demand forecasting AI hasn’t improved accuracy by at least 15%, or your scheduling AI hasn’t reduced labour costs by a measurable amount, pull the plug.

Most importantly, budget for the hidden costs: data preparation, staff training, system integration, and ongoing maintenance. These often dwarf the software costs.

Step 4: Build Internal Capability

The companies succeeding with AI aren’t just buying black boxes—they’re developing internal understanding of how these tools work. Invest in upskilling your team to ask the right questions and interpret AI outputs critically.

Step 5: Plan for Pricing Reality

Current AI pricing models are unsustainable. Build your business case assuming prices will increase significantly over the next 2-3 years. If your ROI calculation falls apart with 3x higher AI costs, the investment probably isn’t worth it.

The Truth About AI’s Future

Here’s what the heated Reddit debate missed: the question isn’t whether AI is a bubble or the next cloud. The real question is whether you’re making smart, measured investments based on operational reality or getting swept up in the hype cycle.

AI will undoubtedly transform how we work, but like cloud computing, the value will come from solving specific problems efficiently, not from having the shiniest tech stack. The companies that will benefit most from AI are those treating it as a tool for operational excellence, not a magic wand for instant transformation.

For product managers in operationally heavy sectors, your advantage lies in being ruthlessly practical. Whilst other industries chase AI moonshots, you can quietly implement proven solutions that deliver measurable improvements to customer experience and operational efficiency.

The AI revolution is real, but it’s not the flashy revolution being sold in boardrooms. It’s the quiet revolution of better demand forecasting, smarter inventory management, and more efficient operations. And that’s exactly where you should be focusing your efforts.