Last week, Jeff Gothelf published a piece that's been making the rounds in product circles: "Storytelling: An AI era survival skill for product managers." The premise is simple but uncomfortable: when everyone in the room can generate data, charts, and research summaries with a quick prompt, the PM who wins isn't the one with the best evidence. It's the one who can tell the best story.
He's right. And it has massive implications for how product teams think about customer insights.
The HiPPO Problem Just Got Worse
Here's a scenario you've probably lived through. You spend weeks doing customer research. You synthesize interviews, identify patterns, build a recommendation. You walk into the room, present your findings, and a senior stakeholder pushes back: "I have data that says something different."
The room goes quiet. Now it's a standoff.
Gothelf calls this the HiPPO problem—the Highest Paid Person's Opinion usually wins. That's always been true. But here's what's changed: that stakeholder's "data" might have taken them twenty minutes to generate. A quick prompt, a synthesized summary, a chart that looks like research.
AI has democratized the appearance of product work. Anyone can produce specs, personas, competitive analysis, and prioritization frameworks. The outputs that used to signal PM expertise are now table stakes.
So if evidence alone doesn't win rooms anymore, what does?
Stories Win Rooms
The difference between a PM who moves a room and a PM who loses the debate isn't the quality of their data. It's the story they've built around it.
AI can generate a research summary. It cannot replicate what happens when a PM says: "I sat with six customers last month, and three of them said the exact same thing without being asked. Here's what that means for this decision."
That specificity—the human presence, the real observation, the judgment call—is the signal that earns trust.
But here's the catch: most product teams aren't actually set up to tell good customer stories. Not because PMs can't communicate. Because they don't have the raw material.
The Customer Insight Gap
Let's be honest about how customer feedback actually works in most organizations:
- Support tickets live in Zendesk
- Feature requests scatter across Notion, Productboard, Intercom, email threads
- Sales calls get recorded but rarely transcribed or analyzed
- NPS surveys collect scores but the open-text responses rot in spreadsheets
- Customer interviews happen, get summarized once, then vanish into a Google Doc
When it's time to build a case for a product decision, PMs scramble to piece together fragments. They remember that one customer said something about onboarding friction—but where was that? Was it in a call? A Slack thread? A churned user's cancellation survey?
The evidence exists. It's just buried, fragmented, and impossible to synthesize quickly.
This is the customer insight gap. And it's the reason so many product debates devolve into opinion battles instead of evidence-based discussions.
Why AI Makes This Problem Urgent
Here's the paradox: AI makes it trivially easy to generate research-looking outputs, but most organizations still struggle to access their actual customer evidence.
Your stakeholder can prompt ChatGPT for "common B2B SaaS onboarding problems" and get a convincing-looking list in thirty seconds. What they can't do is surface what your customers have actually said about your onboarding flow.
That asymmetry is the opportunity.
According to BuildBetter's analysis of AI sentiment tools, the companies pulling ahead aren't just monitoring customer sentiment—they're combining internal team discussions with external customer feedback to spot gaps between what customers are feeling and what teams are discussing. That's the blind spot most organizations miss.
The global sentiment analysis market surpassed $6 billion in 2025 and is projected to grow at 14-15% CAGR through 2030. The race isn't just to collect more feedback. It's to turn that feedback into intelligence that can be accessed instantly, synthesized automatically, and woven into compelling narratives.
Building the Evidence Base for Better Stories
If storytelling is the survival skill, then customer evidence is the raw material. And the PMs who will thrive in the AI era are the ones who solve the evidence problem systematically.
Here's what that looks like in practice:
1. Centralize Everything
Stop letting customer feedback scatter across twelve tools. Whether it's support tickets, feature requests, sales call transcripts, NPS responses, or interview notes—it all needs to flow into one place where it can be searched, tagged, and surfaced on demand.
This isn't just organizational hygiene. It's about building a searchable corpus of customer evidence that you can query when you need to build a case.
2. Let AI Do the Synthesis, Not the Storytelling
This is the key distinction. Use AI to surface patterns, identify themes, cluster similar feedback, and flag sentiment shifts. Let it do the heavy lifting of processing hundreds of data points.
But the story—the narrative that connects the insight to the business context, that makes the room feel the urgency—that's still human work. The synthesis is the starting point, not the deliverable.
3. Capture Specificity
Generic insights don't move rooms. "Customers find onboarding confusing" is forgettable. "Three enterprise customers in the last month have mentioned they couldn't figure out how to invite their team members within the first ten minutes—and two of them said that's when they nearly cancelled" is unforgettable.
The difference is specificity. Build systems that preserve the original customer voice, the exact quotes, the context around when and why something was said.
4. Make Evidence Accessible in the Moment
The best customer insight in the world is useless if you can't access it when you need it. When you're in a meeting and someone challenges your recommendation, you need to be able to pull up supporting evidence immediately—not say "I think there was something about that in a call last month."
This is where AI-powered search and retrieval becomes critical. The ability to ask "What have customers said about our onboarding experience in the last 90 days?" and get synthesized results with links to original sources changes the game entirely.
The Storytelling Framework
Once you have the evidence base, storytelling becomes about translation. The same insight needs different framings for different audiences.
For design reviews: Lead with the user's journey. The friction, the confusion, the unmet need you watched someone struggle with. Make the room feel what the customer felt.
For engineering: Connect the customer problem to technical implications. What's the cost of not solving this? What's the architecture impact of the proposed solution?
For business stakeholders: Frame it in terms they care about. Revenue risk. Retention impact. Competitive positioning. "This onboarding friction correlates with 23% of our churn in the first 30 days" hits different than "users find onboarding confusing."
For executives: Zoom out to strategy. How does this connect to the company's larger bets? What does solving this enable? What's the cost of ignoring it?
The customer insight stays the same. The story changes based on who's in the room.
The PM's Competitive Advantage
Gothelf makes a prediction that I think is exactly right: "The PMs who stand out in the next few years won't be the ones who prompt the best. They'll be the ones who can take everything they've learned—from customers, from data, from experience—and turn it into a story that earns a room's trust and moves people to act."
That's not a soft skill. It's a system.
The system has three parts:
- Capture: Get customer feedback from everywhere into one place
- Synthesize: Use AI to identify patterns and surface relevant evidence
- Translate: Build narratives that connect customer insights to what each audience cares about
The best argument doesn't always win. The best story usually does. And the best stories are built on evidence that's real, specific, and accessible when you need it.
Putting This Into Practice
If your organization's customer feedback is scattered across multiple tools, buried in spreadsheets, or living in individual team members' heads—you're not ready to tell good customer stories. You're not even ready to know what your customers are actually saying.
Start by auditing where customer feedback lives today. Then ask: how quickly can I answer the question "What do our customers think about X?" If the answer is "I'd need a few hours to dig through various tools," you have work to do.
The tools exist now to solve this problem. AI-powered platforms can aggregate feedback from every source, synthesize patterns automatically, and surface relevant insights on demand. The question isn't whether the technology is ready. The question is whether your organization is ready to build the evidence infrastructure that great customer stories require.
Because in a world where everyone can generate research-looking outputs, the PM who can say "I know this because I've talked to customers, and here's exactly what they said" is the PM who wins the room.
That's not something you can prompt your way to. That's something you have to build.