Product managers asking:
“Should I stop focusing on launches and help GTM instead?”
It’s the wrong tradeoff.
If you step away from delivery:
progress slows
you lose leverage with your team
you still don’t fix how the product sells at scale
It's a launch “done” too early.
You didn’t miss your goals this year.
You shipped.
You launched.
You did the work.
But if you’ve sat in on a sales call lately, you’ve probably felt it:
Something didn’t transfer.
The product makes sense to you
…but not to the people trying to sell it.
That gap?
It’s unfinished product work.
Is product return on investment a calculation or a system? The system of behaviors that product teams practice:
💠 Asking hard questions every day
💠 Growing confidence in the hypotheses
💠 Having honest conversations with finance
buff.ly/Z3pXS1M
Early experiments aren’t about proving you’re right.
They’re about finding where you’re wrong—fast.
Each hypothesis carries hidden assumptions:
– who the buyer is
– what would need to change
– what “good enough” looks like
A good experiment surfaces the weakest assumption first.
Pricing estimation for product managers:
How much effort people spend working around a problem.
That effort quietly sets an upper bound on your pricing.
If someone spends ~1 hour/week dealing with it,
your product likely won’t capture more value than that.
A lot of product work stalls.
You build something that works.
Customers respond.
Sales can use it.
It doesn’t fit a feature or a program.
No one quite knows how to move it forward.
Agree! I wrote an article about getting to a hypothesis for startup PMs that inspired this post 😀 open.substack.com/pub/amycmitc...
At some point, more discovery stops adding clarity
and starts delaying decisions.
You don’t go from learning → certainty
You go from learning → hypotheses
Pick 2–3 plausible paths.
Make them explicit.
Then try to break them.
Progress comes from ruling things out faster.
Early-stage product managers rely on instincts:
→ Follow the friction people live with
→ Follow the workarounds they tolerate
→ Follow the decisions they struggle to make
Money comes later.
This is how you earn the right to see it.
How to connect strategy to product work.
- connect your core assets to customer value
- connect your core processes to customer value
Prioritize strategic projects:
1 customer value
2 baseline requirements for customers
3 postpone the rest until 1 and 2
buff.ly/X4W1xDA
Small inputs from product can ripple outward:
Engineering aligns earlier around constraints and tradeoffs
GTM shapes clearer positioning
Sales builds confidence in conversations
Clarity shared early tends to compound over time.
You don’t need a full strategy to be useful.
A few inputs at the right moment can go a long way:
– customer trends
– competitive signals
– emerging use cases
Shared early, they help teams move forward with clarity.
updates don’t move complex work forward
they expand it
the real job is getting it to converge over time—with leadership involved
The most valuable PM move is reducing the distance between:
customer truth → product decisions
When that distance shrinks:
alignment improves
decisions come faster
teams move with more confidence
It changes the shape of the work around you.
Delivery teams plan around architecture, scale, and constraints.
What they need from product is clarity grounded in real customers:
how usage is evolving, where demand is heading, what actually matters.
Even rough notes can unlock better planning.
What availability means for product managers:
🔹 Making time for crucial people and decisions
🔹 Making space to be fully present
🔹 Making calls by deciding in a timely fashion
🔹 Making good by following through on commitments
buff.ly/1EBXbdH
You don’t need a big system to improve product context.
Start with 5 things in one place:
product + service descriptions
pricing + packaging
release notes
sales FAQ
AI context
Then add just enough structure and a simple cadence.
When your context isn’t maintained, something subtle happens:
Your product stops being a system of truth…
and becomes a system of interpretations.
Sales says one thing.
Docs say another.
AI says something confidently wrong.
Alignment doesn’t break loudly.
It erodes quietly.
When you get this response to your product:
"What's the full solution?"
It usually means your product is hitting an adoption wall.
I wrote about building repeatable solution packages that are easy to buy:
Product knowledge isn’t documentation anymore.
It’s how your product behaves when you’re not in the room.
Sales calls.
Support tickets.
AI copilots.
If your context is messy, your product doesn’t scale.
Product managers who treat this as a growth lever (not admin work) move faster.
A “quick question” is rarely quick.
“Does analytics come with the base SKU?”
→ 45 minutes later you’ve updated pricing, FAQs, and your AI context.
That’s product knowledge drift showing up as real cost.
It's a product context maintenance problem.
table of product manager beginnings for developing cognitive scaffolds
The question about how beginners can build their cognitive scaffold with or without AI is a good one. I put some definitions of "beginners" in the product management context in a quick table to think about the question. What do you think about this for a starting point?
I'm using AI to manage that cognitive load, not to skip the thinking. It organizes the 'table stakes' facts so I can spend 100% of my energy on the tradeoffs and unwritten rules. Do you think there's a threshold where data volume makes 'manual' retrieval a net-negative for learning?
I hear you on the scaffolding—that's a fair challenge. My view is that the sheer volume of data can sometimes exceed our bandwidth, burying the 'aha' moments in noise.
We are all beginners at some point in time when learning something new. When we are beginners (or early career), we can get to a point of view by learning faster with AI. And with that POV, we are ready to drive tradeoffs and outcomes faster.
I believe the valuable parts of product management are making tradeoffs and being accountable for outcomes. To effectively do this, you need to quickly discover and consolidate multiple data points.
I'm finding that AI helps with learning, so I can add my experience for the planned outcome.
You are right that learning comes from retrieval. The challenge is summed up:
"The organizations that figure out how to develop judgment in early-career workers, not through years of retrieval grind but through structured exposure to decision-making, will have a massive talent advantage."
I'm suggesting that product managers use AI for retrieval and synthesis to free up time for:
Judgement
- Evaluating options with context AI doesn't have.
- Weighing trade-offs that involve relationships and unwritten rules.
Decision
- Putting your name on the outcome.
Structure for using AI to free your time for valuable product work:
Layer 1: Retrieval
Layer 2: Synthesis
Layer 3: Judgement
Layer 4: Decision
"Use AI aggressively at Layers 1 and 2, collaboratively at Layer 3, and keep Layer 4 firmly human."
buff.ly/wkHheYc
Momentum in product work comes from small moves that teach you something.
3 emails instead of 30.
1 quick test instead of a full rollout.
A partial answer instead of waiting for certainty.
You don’t need more time or permission.
You need a step small enough to learn from now.