4 Principles of Upskilling for the New-Age PM
How to think about skill development when uncertainty is the only constant.
In my article 3 Seismic Shifts in Product Management I explained the main shifts reshaping tech (and the world).
Today I want to explore what these shifts mean for the skills Product Managers need to develop - skills that are no longer optional.
Shift to consumerized work software
This shift brings B2B PMs closer to individual users and demands enhanced customer knowledge (buyer and user level), data-driven behavior analysis, understanding growth pathways, and self-serve journeys. Sounds duh, but collectively we’re still not great at this.
Shift to efficiency-first mindset
Simply put, PMs must now justify their employment with numbers sitting as close to profit as possible. Every initiative should indicate how it contributes to business health and prove estimated impact.
Shift to AI-enhanced workflows
This has direct and indirect effects:
Direct: Non-critical tasks get automated. AI speeds PMs through information processing, project management, and mundane tasks across the value chain. As Ravi Mehta noted in his recent article, "Product Managers will do more true product management." But as tactical work disappears from calendars, critical skill gaps become visible.
Indirect: Delivery is faster. AI made engineers way faster, but they're now blocked by PM decision speed.
Competition is fiercer too - you build faster, but so do competitors. Meaningful innovation, clear differentiation, and seamless experience are required just to stay afloat.
This creates a new-age PM who must lean into their original mandate as business owner while becoming more effective in practice. PMs need to step up their game in innovation, impact, velocity, and leadership. For some of us this will mean lots of new skills to learn, for others less.
Here I want to share a mental model of how to think about upskilling. So to say, teach you how to fish. Here are four principles for new-age PMs:
Become the true mini-CEO of your product (yes, finally!)
Let's dive in!
Become the Mini-CEO
The product community has long rolled their eyes at the "mini-CEO" concept (usually out of bitterness). But the table is turning, and the time is coming. Companies are finally feeling the need for PMs who can actually operate at this level. Time to stop laughing and start learning.
If you want to understand what "mini-CEO" actually means, just observe your real CEO. They know their product, market, technology, and customers intimately. They make decisions quickly. They understand how different functions contribute to the whole. They're responsible for outcomes even when they don't directly control all the inputs.
To help you become that, it is crucial to develop a few skills though.
Commercial fluency
I’m not saying become a CFO overnight and own financial forecasts. But you must know your business profit and revenue structure, key commercial and business health metrics, and how your work connects to dollars.
Every initiative should pull at least one of the profit levers:
Increase revenue
Protect revenue
Save costs
Avoid costs
This isn’t nice-to-have knowledge. It changes how you communicate, plan, prioritize, and structure your work. Before it was authority, guts, and random data. Now you have a unified language to talk about opportunities whether they come from CEO, Sales, customer interviews, or your crumbling unit economics.
It also gives you a new perspective on your and your team’s work: now you won't spend weeks and thousands on poorly structured research just because “this is just how we work”.
And when VC playbooks are breaking and uncertainty is through the roof, it isn’t just helpful - it’s protection: execs keep PMs who drive business value, not just ship features.
There's a sad-but-true article from Leah explaining why no one is actually safe in this environment: The 250k per Employee cultural KPI causing layoffs in tech
I’ll be writing more both on commercial skills for PMs, and on running and sizing discovery projects, so make sure to subscribe to not miss these posts :)
Growth & GTM fluency
Growth is everyone’s business. Sure, there are specialized Growth PMs chasing metrics performance, but that doesn't mean the rest of us get to ignore it. Every PM needs to become growth-fluent to some degree.
Here’s what this means:
GTM basics: How does your product's value actually reach customers? What are the implications for your planning and delivery?
Build Value Ecosystems: Shipping features is not your job. Your job is to ship value and make sure customers find it, adopt, engage with it, and pay for it. And this should be baked into your planning and development process.
Understand pricing: Most PMs have zero clue how pricing models impact growth or influence value development (I got you! 5 JTBD of your pricing model). This matters even more now with usage-based pricing on the rise and AI products that actually cost money to run.
Know your buyers: You might know how users engage with your features, but do you understand how purchasing happens? What economic buyers care about? How do sales and procurement cycles work?
End-to-end customer perspective
The bigger the company, the more PMs get trapped in their isolated feature bubbles. They become experts at their corner but have no clue what happens before or after their handoff.
This is backwards. CEOs don't run companies by only knowing their department — they understand how everything connects, even if not deeply everywhere.
PMs need the same breadth. You need to understand how they navigate the entire product, what touchpoints they hit, where they transition between flows, where they fail, customer account dynamics, and how this all varies across customer segments.
Why? Because your customers don't have “engineering problems” or “design problems” or “billing problems”— they have problems with your product. Their struggles can surface anywhere in their lifecycle and journey, and if you're wearing blinders, you'll miss the opportunities in the gaps between teams and touchpoints.
Cross-functional collaboration, not just facilitation
I am triggered by the word “facilitate” because often it means just organizing meetings and leading workshops. Real collaboration doesn’t happen in your calendar.
As I just mentioned, opportunities live in cross-functional gaps. Customer and market knowledge sits in sales, marketing, CS, even legal. To find and capture these opportunities, you need to work with these people as actual partners. And you need to understand how they work, their goals, processes, reasons, challenges, etc. You can’t collaborate with people you don’t understand. Maybe it’s just me, but I’ve often sensed some implicit war between “tech” and “business” departments. This is unproductive and kills innovation.
True cross-functional collaboration means working together (not just with each other) toward joint objectives — not only meetings where you review Jira tickets together.
Organizational leadership & influence
Most PMs are terrible at executive communication. They get stuck in feature-level conversations while leadership cares about impact and risks.
The new reality demands more senior PMs — not in titles (those will likely deflate), but in capability. PMs who can co-create strategy, rally cross-functional peers around shared goals, and get buy-in by being genuine business partners who understand what the company, leadership, and market actually need.
The Five Pillars of PM Influence
Speak their language: Execs think in dollars, not features. They care about customer impact, revenue impact, and vision impact. Learn the commercials.
Master alignment: The more you understand your peers' goals, incentives, and daily struggles, the more you can craft proposals that help them win too. When people see you're solving their problems, not just pushing your agenda, influence becomes natural.
Build trust: Trust has two components: intent and reliability. Intent means people believe you're genuinely trying to help the business, not just advance your career. Reliability means you keep people in the loop, follow through on commitments, and if things go sideways — you say so.
Establish credibility through data: A leader's most important job is building conviction. Charisma helps, but solid data is the universal trust driver. Do your homework. Gather evidence. Present clear pointers that prove both your impact potential and the credibility of your ideas. Half-baked proposals with gut feelings don't cut it at the exec level. I was that person, so I know.
Perfect your narrative: Build clear, logical storytelling that resonates with stakeholders and gets them excited to help you. Your narrative should connect the dots between customer problems, business opportunities, and proposed solutions in a way that makes their buy-in a no brainer.
Executive Communication: It’s the other way
Most PMs communicate like they're writing a novel: setup, struggle, journey, resolution. Execs want the exact opposite.
Lead with the essence: problem, insight, recommendation, necessary resources. Dive into details on request (but have them handy!) CEOs don't care about your research methodology or every customer quote — they need enough context to guide their questioning and decision-making.
Think executive summaries, not hero journeys. Get comfortable with the discomfort of cutting straight to the point.
Expectations Management & Responsibility
Keeping in touch with reality is a universally underappreciated skill
We're chronically bad at this. We overpromise on timelines and impact not from malice, but from bias. We want to impress, so we throw around great-looking numbers that feel good in the moment but destroy trust over time.
Here's a financial planning trick that I stole. Create three scenarios: pessimistic, realistic, optimistic. You don't need to present all three, but thinking across them keeps you grounded and demonstrates thoughtfulness.
And let's talk responsibility. Yes, you rarely own outcomes entirely — that's organizational reality. But you absolutely own the outputs that contribute to those outcomes. Responsibility means owning your shit. That’s how we learn and gain trust.
Remove Yourself from the Equation
I spent years trying to figure out what made certain execs exceptional. Then in one performance review, my manager said: "Alex, learn to remove yourself from the problem-solving equation. There is no “I” there".
That was it. These leaders provided exact solutions because they didn't filter decisions through "how does this affect me?" The problem existed, the solution was clear, so it got done.
When your ego enters the equation, stakes get personal. Personal stakes cloud judgment. You optimize for looking good instead of doing what’s right. You avoid uncomfortable conversations. You delay decisions that might reflect poorly on past choices.
Self-removal is rare because it's unnatural. But it generates unbiased solutions and, paradoxically, makes you more influential. People trust leaders who prioritize the right answer over their own comfort.
Two practical approaches:
Get an outside perspective. Find a coach or mentor who cares about your success but doesn't share your stakes. They can see risks you're blind to when you're inside the problem.
The "Frank" method. Write the situation factually — no adjectives, no emotional language. Assign it a random name: "So Frank faces situation XYZ. Here's what I'd tell Frank to do." This reframing creates enough distance to see clearly.
TL;DR: The nasty part of me would say “it’s time to grow up”. The professional one says “get real and learn to partner with your leaders, even if it is uncomfortable and even seems risky”.
Become Antifragile
Today, uncertainty is unprecedented. Pandemics, wars, economic upheaval on one side, and overwhelming pace of innovation and change on the other. There's barely any footing anymore. Old ways are crumbling, new ways are in the making, and predictability has become an illusion.
Becoming antifragile means embracing this uncertainty and actually thriving because of it. At its core lies incrementality — the ability to make small, reversible moves that compound into strength over time. While big bets can break you when they're wrong, incremental progress lets you adapt, learn, and actually benefit from volatility.
Today, this isn't just philosophy — it has to be operationalized.
Adaptive Planning
I heard this term recently and loved it because it reflects our reality. While not foreign to product management, it was always discussed in "should" terms but rarely practiced. Today it must be the status quo.
Outcome-driven planning
If you're still shipping features, you're confusing means with ends. The job is to build value ecosystems, and value isn't some static unit like a user story — it’s measured by impact on customer and business outcomes. So why not focus on them? Be rigid on outcomes, flexible on how to reach them.
You may be rolling your eyes "what are we, 2020?" If so — good for you. But lots of teams create annual feature-centric roadmaps without much regard for the outcomes they intend to bring. Why? Organizational legacy, resistance, perception of the product team’s job to “just focus on customers and ship features”, and even their inability to impact-size maintenance or tech-debt work.
We're not in the "invest money to build" business anymore — we're in the "invest money to make money" business.
Incremental success
Launches rarely succeed overnight. They need time and optimization to reach potential, and this time must be accounted for. Plan for iterative development, even though it's uncomfortable — you can't see far ahead and must rely on continuous learning to steer forward. The mantra is "smaller, faster, tighter".
Your business still may be planning in quarters, but you need to build your cycles on much shorter timelines: monthly, bi-weekly, or even weekly.
Built-in accountability
Customer and economic outcomes should be built into team charters and org design. This doesn't mean only one team owns a metric — it means every team has clear accountability for specific outcomes that ladder up to company goals.
But when teams are accountable for outcomes (not outputs), they need to work together differently. Everyone's success becomes interdependent. The team owning the "AOV" metric might need another team's support to change core product functionality that sits in their surface area.
This creates operational “chaos” as teams must collaborate across boundaries and factor partnerships into their planning. This gets exponentially harder when every team has a dozen OKRs. The solution: shift to smaller, more focused goals and give teams autonomy to figure out how to work together to reach them through adaptive, incremental development.
Experimentation
In this setting, you must reduce uncertainty with minimal time and effort.
PMs are no strangers to “discovery”, but very often it ends with forming hypotheses from customer interviews and data. This does not reduce uncertainty — the Build-Measure-Learn loop does. This is experimentation, and it should be baked into your discovery process.
Experimentation is often equated with A/B testing only. In reality, it includes prototyping, smoke tests, surveys, pre-sales, and more. When teams lack traffic, fast setup, or clear data, they often resort to nothing. But every day a hypothesis remains untested has an opportunity cost.
The objective is finding the best way to sense-check your solution with the market, appropriate for your development stage and resources. And it doesn’t have to be perfect or definitive — it has to be directive. So the question isn’t "how do we run an A/B test in our B2B product that lacks traffic?" It's "what's the fastest valid experiment we can run right now?"
With today's rapid prototyping tools, it's about adopting the mindset and operationalizing it into your process, not running isolated experiments to showcase best practices. Experimentation is an operating system of innovation.
TL;DR: Uncertainty isn't going away: make small bets, test fast, fail cheap, and build capabilities that thrive on chaos instead of crumble under it. This is yet another way in which product teams need to work like growth teams.
Don’t Be The Bottleneck
Previously, squads had 4-7 engineers, a designer, and a PM. It was fair — building took forever, and PMs had time to plan their next work. No more.
AI has made engineers way faster. The bottleneck shifted from building and coding to making decisions about what to build next. It’s a new unexpected challenge for PMs: match this engineering velocity or become the constraint that slows everything down. So no, I don't buy the "AI will replace PMs" hype. AI can't replace nuanced judgment, empathy, and creativity — but it demands we work differently.
Proactive Opportunity Management
PMs need to stay ahead of the curve in generating and assessing opportunities. This means working with insights proactively. There are multiple sources for these insights: impact assessment of releases, data analysis, costs and unit economics improvements, customer feedback, market research, and more.
We've always done this work. The difference? It was reactive:
Manual investigation when conversion rate drops
Customer interviews because we're planning feature XYZ for Q3
Digging through customer request backlogs to validate ideas
This won't cut it anymore. You need to get ahead with:
Proactive data analysis: Slice and dice product and behavior data to find correlations, outliers, inconsistencies. Product analysts partnering with PMs, designers, and engineers to proactively hunt for insights and co-own improvements isn't optional anymore.
Continuous customer conversations: Talk to customers for no reason at all. Some of my best insights came from general conversations that weren't part of formal user research.
Pulse monitoring: Keep tabs on other teams' releases, cross-functional reports, market trends, competitor moves, industry experiments.
Signal collection: Create channels where employees can send insights and proposals. Make it easy for the organization to feed you intelligence.
But insight generation is only half the equation. You also need to get better at continuous impact-sizing and business-casing ideas, then feeding them into your experimentation pipeline.
The main shift: this work used to happen in the context of specific initiatives to prepare for delivery cycles. Now it should be decoupled from roadmap initiatives and done continuously. By the time engineers finish their current work, you need a backlog not of ideas, but of validated opportunities worth investing in.
Operational Velocity
I'm deliberately not creating the "AI skills" category. We're still early in this timeline — things are swirling and forming, and it'll be a few years before we can look back and understand what really mattered. Some PMs work on AI features and need technical depth. Others don't.
Everyone wants to seem like they've figured it out, but I see far more discussion about running companies with 50 AI agents than companies actually doing it. The pressure to try everything, build endless automations, and completely reimagine your workflow is FOMO at this point.
Yes, increased engineering velocity is real. But it's not always dramatic, and organizational inertia takes time — especially if you're not a shiny new AI-native startup (and if you are, you don't need my advice).
If you're naturally curious about AI, experiment away. But if you're feeling outside pressure to build agents for everything while juggling a full workload? Take it slow. Don't let anyone shame you for not being obsessed with every new tool. Your natural interests don't have to go there, and that's completely normal. Here, I said it.
I'm not saying do nothing. I'm saying if you're coping with your current workflow, or if your organization is stuck in old ways, there's time to learn without anxiety. Start with simple improvements you probably already make in your life and work. Focus on what you actually need today, not imaginary agents for problems you don't have.
For specific ways to improve PM workflows with AI, I recommend a brilliant recent article from Ravi Mehta How AI is supercharging every phase of the PM lifecycle. Let that be your guide.
Decision Making
Decision-making isn't about having all the information — it's about understanding what's at stake: impact and (ir)reversibility.
Strategy requires high conviction because you're committing to one direction long enough to go through cycles of failure and learning before it works. Everyone needs to believe in it and channel their energy into it. So be robust here — gather data, analyze carefully, make this one thing impeccable.
Tactical decisions are different. They're about finding the right path to make your strategy work, fast. Here, the cost of indecision usually exceeds the cost of being wrong. Every day you delay is an opportunity cost, and a lack of decision is still a decision btw. Your job is to lower the stakes wherever possible: make decisions reversible, reduce blast radius (remember: small, fast, tight). And normalize failure to learn.
Teach your team judgment and create decision frameworks so they stop escalating everything.
Watch for biases that cloud thinking under pressure - confirmation bias, sunk cost fallacy, anchoring.
And remove yourself from the equation when stakes get personal.
TL;DR: Engineering velocity skyrocketed. Product decision-making didn't. Close the gap with being proactive, improving your efficiency, and learning to make fast decisions.
Own Your Game
At the end of the day, it's all about you. The only person who will define your success is looking at you in the mirror.
It's hard times, a hard world, and a hard job — and you need to be up for it. I'm not talking the hustle-bro-neostoic culture. I'm talking about setting yourself up for long-term sustainable success through inevitable highs and lows.
Vertical expertise and positioning
In fierce competition, niching becomes your differentiation. For businesses and for the workforce. Your niche can be big, but make it concrete. Rally your focus toward building this expertise while maintaining general fluency in other areas.
Most often this takes the form of domain specialization. Although industry knowledge doesn't necessarily translate into great functional performance, great functional performance supplemented by domain expertise wins in this market. Or go functional — Technical PM, Growth PM, AI PM, Marketplace PM…
Learn to market yourself
I’m not necessarily talking about personal branding, though it helps. But more importantly, market yourself internally. You don’t have to be a shameless self-promoter or detached from reality. But if you don't tell people about your achievements, you won't get deserved recognition. It's your job to communicate your impact and value, not anyone else's.
Be strategic about everything
From product initiatives to career to life choices. You don't need to live through spreadsheets, but be deliberate and keep the main things the main things. Early in your career, flowing with life is normal. But after a few years, take ownership of direction, or a crisis is guaranteed.
Carve out "forest time" — strategically assess goals, progress, direction, and course-correct. Weekly or bi-weekly for work, quarterly for life.
Learn to restore
I don't believe in work-life balance. Not for those seeking mastery. Mastery is sacrifice. But I recently realized how crucial true restoration is — not just travel, dinner with friends, or Netflix.
This pace of change and information processing has profound implications on our nervous system. We weren't built for constant fight-or-flight mode. No wonder mental health deteriorates despite medical breakthroughs. Anxiety, irritability, focus difficulty, inability to relax, overwhelm — these aren't normal.
An evening off isn't enough, and weekends tend to drown in errands. We need time-tested tools: detachment, digital detox, mental space, exercise, breathing, connecting with nature. Make time for this.
Understand yourself
Sooner or later you'll understand that being happy means being yourself — nothing less, and nothing more!
Compromising your identity, values, goals, principles to fit someone else's image of how a PM (or whoever) should be, is ridiculous and not worth it. What gives you fulfillment is leveraging your individual traits and strengths, and following your interests — not numbers on someone’s spreadsheet.
Learn about yourself early. What follows is a long period of integrating this knowledge into life and work, reconciling who you are with what the world wants you to be.
Build your support circle
In the end, you're alone — more so in this job.
For some, connecting is easy. For others, not so much. If you're in the second group, find a support circle of friends, professionals, mentors who are or were in your shoes. To learn, share, vent. Even if you don't think you need it, try it — it feels good. There's no decree that you must keep everything inside. Let it out, cleanse a little. You'll be lighter.
TL;DR: Niche down, speak up about your wins, stay strategic, and learn to truly detach. This job is hard and lonely, so build the personal foundation that sustains you through the neverending cycle of highs and lows.
What about leadership?
It's not entirely clear to me yet. I've spoken to quite a few product leaders in the past months, and while there's general agreement that "the core of the job stays the same," there's significant volatility around what new skills and approaches they should master. Remains to be seen.
Overall, I think the job will skew in the directions described above, just at a higher level:
Product strategy will align more tightly with growth and business strategy.
Product leaders will increasingly own P&L, team efficiency, and participate in annual financial forecasts.
They'll spearhead cultural and operational changes around antifragility, growth mindset, new collaboration models, autonomy, and AI-enabled workflows.
And of course, even more emphasis on coaching — both unchanging fundamentals (more critical than ever) and new skills and mindsets.
Good luck!