AI – A Boon or Bane for Product Managers?

Are you a product manager with FOMO when it comes to Artificial Intelligence? Should you hop on the AI bandwagon? AI has become the buzzword for organizations, big and small, because it carries an appeal and adds a touch of innovation.

Try this with me:

Here’s presenting to you our latest “AI-powered Smart Watch” vs Here’s presenting to you our latest “Smart Watch edition”

Didn’t the former sound more appealing and better all? Even though you might not fully understand what AI-powered means! That’s the power of AI.

The truth is, a smartwatch might have some intelligence built in (after all, why call it smart?), but just mentioning "AI" makes it sound cooler—even if there’s no real AI component.

Are companies truly leveraging AI’s potential, or are they just riding the wave by slapping the AI tag on everything?

You might ask, "What does all this have to do with product management?" Well, I had to get your attention, didn’t I? Now that I have it, let’s dive in.

Much has been said about product management, and much has been said about AI. In this blog, I aim to explore different perspectives on how AI impacts product management.

Did you know that by 2025, 75% of product managers are expected to use AI tools for decision-making?

Let’s discuss five ways AI can benefit product managers (starting with the positives, of course!):

AI for Deep Understanding of the Customer

They say, "The customer is king!" So why not start with understanding what the king wants? Defining your customer, who has many attributes (e.g., demographics, psychographics) across various segments, can be overwhelming. But with AI, you can map target customers into clusters, which helps you define 2-3 target personas more effectively. This provides a deeper understanding and helps you serve them better.
AI for Competitor Analysis

Once you understand your customer, it makes sense to understand your competitors. Conducting a detailed competitor analysis takes time and resources. Thanks to AI, you can automate this by defining parameters for ranking your competitors, and it will generate the analysis for you! It’s not only easier but also super-efficient.
AI for Prioritization

With insights into customers and competitors, how do you prioritize different features? Any product manager knows how challenging it is to prioritize a roadmap—balancing customer requests, stakeholder alignment, and technical constraints. Predictive analytics can help prioritize tasks based on data and assigned values, not just gut feelings or falling into the HIPPO (Highest Paid Person's Opinion) trap.
AI for Collaboration

After addressing the first three points, it’s time to collaborate with teams to execute plans. AI enhances communication through tools like Slack (e.g., reminders in channels, Zoom integration for ease of use), making day-to-day tasks more manageable. Zoom’s ability to summarize discussions and provide updates to those who couldn't attend is remarkable.
AI for Analyzing Customer Feedback

Any product relies on customer feedback for growth. AI can quickly analyze hundreds of entries to identify common frustrations and sentiments. Using NLP, it can understand emotions and categorize feedback as positive, negative, or neutral.

Now, let’s explore five ways AI can be challenging for product managers (yes, it’s time for the ‘bane’ side):

AI, Not Understanding Sentiments

From personal experience, I’ve received product recommendations that weren’t exactly private-friendly. When using shared family accounts, AI should be able to interpret sentiments and avoid awkward recommendations. It has a long way to go in this area.
AI, Difficult to Explain Benefits

As a product manager, you need to explain new features to users, but what if your team’s development involved improving a complex ML model? For example, how do you explain that a "Boost Model" has been replaced with a "Perk Model" (just made-up names) and make it sound exciting in a feature release? It’s challenging to convey the value of intangible improvements.
AI, Not Always Accurate

I might get some backlash for this, but most LLM models have a maximum accuracy of 95%, which means they cater to the majority. But what about the remaining 5% of your users? If you’re in that minority and face issues, it can be a problem. AI tools need to improve their ‘confidence’ metric.
AI, Not Easy for Adoption

With all the AI features available, adoption can be tough, especially for users like senior citizens. As product managers, it’s crucial to understand your audience and focus on adoption rather than just implementing AI for its own sake.
AI, Not for Everything

Last but not least, don’t add AI to all your features just because it sounds cool (contrary to how I started this blog post). While it might make sense for certain cases, don’t generalize—especially if it makes it difficult to explain what the "AI" component actually does.

Now coming to decision making of whether it is a boon or bane for product managers. I would say AI is inevitable and is here to stay. Use it to your benefit, but also be mindful of its challenges, so you know where to apply and where not to. Love it or hate it, but you can’t ignore AI – understanding different perspectives will help tilt needles in your product decisions.