Episode 14: How a Wrong ICP Breaks Your Product or Service: Why You Keep Building Features Nobody Asked For

Episode: 14
📅 Published: May 15, 2026
🕐 Duration: 20:16
Episode Description:
You built something real. It works. Customers are using it. Just not quite the way you built it to be used.
That gap is one of the most expensive product positioning problems in marketing for entrepreneurs — and one of the hardest to see while it’s happening. The product is shipping. Revenue is coming in. The team believes in the vision. What’s underneath it, usually, is a set of assumptions about your customer that were never validated against the people actually buying.
This episode looks at what happens when a wrong ICP gets locked into a product or service — through engineering time, design decisions, and roadmaps built on conviction instead of evidence. The Amazon Fire Phone and why it failed (it wasn’t the wrong customer — it was wrong assumptions about what that customer would trust Amazon with). Google Glass, and the enterprise customers who raised their hand before anyone thought to ask. And a concept from urban design called desire paths that reframes what customer workarounds are actually telling you about your market positioning.
What this episode covers
- Why wrong product failures are the most expensive ICP problem — and why they’re invisible while they’re happening
- The difference between Apple’s ecosystem trust and Amazon’s commerce trust — and why it matters for entrepreneur marketing strategy
- Vision without validation: how product roadmaps get defended by conviction instead of corrected by evidence
- Desire paths — what customer workarounds are actually telling you about where your product positioning should have been
- Google Glass and self-emerging validation: the right customer showed up before anyone thought to look
- How to talk to customers who stayed and customers who left — and what each conversation can and cannot tell you
Resources
- Brand Therapy — if your product or service feels misaligned with who’s actually buying: https://greyleafmedia.com/diagnostic
- Free ICP Toolkit (15 pages) — walks you through exactly who’s actually buying from you and how to build your messaging around them: https://greyleafmedia.com/find-your-icp
- WordPress design, development, and hosting: https://greyleafmedia.com/services
- More episodes: https://greyleafmedia.com/podcast
- Connect on LinkedIn: Jason Haeger
Read the transcript
I'm Jason Hager, and this is marketing notes for entrepreneurs from Grey Leaf Media,
where impact driven leaders use trust, clarity, and strategy to turn good work into sustainable
growth.
This is the sixth episode in our series on how a wrong ICP cascades through your entire
marketing system.
Last time, we talked about how a wrong ICP breaks your sales messaging.
Today, how it breaks your product or service.
Every other failure in this series is recoverable relatively quickly.
Wrong social media strategy?
You repost to the right platform.
Wrong content?
You create different content.
Wrong positioning?
You rewrite the homepage.
Wrong sales messaging?
You change the conversation.
Wrong product is different.
A product built on wrong assumptions about your customer locks months or years of development
into the wrong direction.
Engineering time, design decisions, infrastructure choices, all made in service of a customer
you thought you understood but didn't.
A service built the same way carries operational debt, delivery structures, team skills, pricing
models, onboarding processes, all shaped by assumption that were never validated against
the actual human buying.
This is the most expensive manifestation of a wrong ICP and it's the hardest to see while
it's happening because the product is real.
The service is being delivered.
Customers are using it.
Just not in the way you built it to be used.
This is episode 14, how a wrong ICP breaks your product or service.
Let's talk about building on wrong assumptions and what it costs.
Building on wrong assumptions.
In 2013, Amazon launched the Fire Phone.
The product made a certain kind of sense from the inside.
Amazon's most engaged customers trusted the platform deeply for commerce.
Firefly, a feature that let you point your camera at almost anything and buy it on Amazon
instantly, was built for this customer.
By perspective, a 3D display effect driven by headtracking was built to make the device feel
premium and differentiated.
The features were real.
The engineering was sophisticated.
The product worked.
But the assumptions about what the customer would trust Amazon with were wrong in a specific
way.
Apple earns a particular kind of trust.
They earn it through privacy, through products that work the way they say they will, through
a business model that makes money by creating things people genuinely want to pay for.
That consistency across every device, every service, every year is why Apple customers follow
them into every area of their lives.
This trust built brick by brick over decades of fulfilled promises.
Amazon earns a different kind of trust, real trust, but for a specific thing.
Competitive prices, fast delivery, a return process that works.
That trust is meaningful, but it's built on a different foundation because it's built
on a different business model.
The fire phone asked Amazon customers to hand their daily digital life to a platform
whose relationship with them was fundamentally transactional.
The ICP existed.
The assumptions about how far that trust would extend were wrong, and the entire product
was built around those assumptions.
The fire phone was discontinued 14 months after launch.
Not because the customer was a fiction, because the assumptions about them were.
The same pattern plays out in service businesses without the headlines.
A firm structures its packages around what it believes clients need based on untested
assumptions.
The deliverables make sense on paper.
The methodology is sound, but actual clients keep asking for something adjacent, something
the firm doesn't offer, or offers as an afterthought, because it didn't fit the assumed
engagement model.
The firm keeps selling the package.
Clients adapt around it.
Everyone's slightly frustrated, and nobody names why.
The three ways it breaks.
When your product or service is built on wrong ICP assumptions, it breaks in three specific
ways.
First, you build around needs that aren't actually central.
Usage data looks fine on the surface.
People are engaging.
Sessions are happening.
Deliverables are being delivered.
That look closer.
Which features are actually being used?
Which parts of your service are clients engaging with deeply?
And which ones are being quietly ignored, worked around, or politely tolerated?
When features go unused, the default assumption is that customers don't understand them.
So you invest in education, better onboarding, clearer documentation, more thorough orientation.
But sometimes the feature isn't misunderstood.
It's actually irrelevant.
It was built around an assumed need that your customer doesn't have.
And your roadmap starts protecting the vision, instead of responding to reality.
Every product and service has a vision.
That vision is usually built around your assumed ICP.
What they need, where they're headed.
What would make the product significantly more valuable to them?
But actual customers are asking for something different.
Something that keeps getting deprioritized because it doesn't fit the vision.
Something that keeps showing up in support tickets, in sales conversations, in off-hand comments
during onboarding.
That's on the roadmap.
Great feedback.
And then, nothing changes.
As the roadmap was committed before that feedback existed.
Vision without validation is how you spend two years building something your actual customers
don't need.
Third, customers quietly repurpose the product into something adjacent to what you designed.
This is the quietest failure.
Users don't complain.
They find workarounds.
They use your product or service for something close to what you built it for, but not quite.
They're getting value.
Just not the value you designed for, and not at the depth you anticipated.
Retention looks acceptable.
NPS is middling.
Nothing is obviously broken.
But when you look at your best customers, the ones who stay actively, refer genuinely,
and get the most from what you offer, they're often using the product differently than
everyone else.
They've found a fit you didn't explicitly design for.
The product is working for them, not because of your assumptions, but despite them.
That's information.
Most companies file it under customer stories and move on.
Why it's hard to see.
In 2013, Google launched Glass as a consumer product.
The ICP was the connected tech-forward early adopter who wanted digital information layered
seamlessly over daily life, in public, in social settings, as part of how they moved through
the world.
The consumer ICP never really materialized.
The social friction was significant, wearing a camera on your face in public created resistance
that no feature refinement could resolve.
But something else happened.
Surgeon started using it to review patient data during procedures without looking away from
the operating field.
Warehouse workers used it to access inventory information hands-free.
Feel the technicians followed repair instructions without putting down tools.
These customers weren't in the plan.
They weren't the ICP.
Nobody had a product roadmap built around them.
But they were there, using the product, finding genuine value, and asking for more.
What Glass needed wasn't a better consumer product.
It needed to get out of the way and follow this signal that was already there.
Google eventually did.
Glass Enterprise Edition launched in 2017.
The product that failed publicly as a consumer device became a legitimate tool in controlled
professional environments.
For a customer that had essentially raised their hand and said, "This is for me," before
anyone thought to ask.
The lesson isn't just that Google pivoted.
It's that the right customers self-identified before the company was paying close enough attention
to notice.
The validation didn't come from research.
It came from reality insisting on itself.
This is why wrong ICP product failures are hard to diagnose.
Because some consumers are using the product, revenue is coming in.
The team believes in the vision.
And belief in vision is exactly what makes wrong assumptions durable in product development.
Feedback that doesn't fit the model gets deprioritized.
Roadmaps get defended by conviction.
By the time the gap becomes undeniable, significant time and capital have already gone into the wrong
direction.
How to diagnose it?
Or questions worth sitting with, honestly.
What features or service components do customers consistently ignore?
Not used incorrectly.
Ignore.
If something you built rarely gets touched, ask why before assuming it's an education problem.
The answer might be simpler and harder.
It's not solving a problem they actually have.
What workarounds have your customers built?
There's a concept in urban design called a desire path.
It's the trail that forms when enough people walk somewhere unofficial, cutting across a lawn,
taking a shortcut through a gap, because the formal path doesn't go where they actually
need to go.
Some planners now deliberately wait to see where desire paths emerge before deciding where
to lay the pavement.
The path tells them where people actually need to go, not where someone assumed they would.
Product workarounds are desire paths.
When a customer builds a workaround, they're showing you where the pavement should have been.
The workaround isn't a failure of adoption, it's information, map it.
What do customers ask for that surprises you?
Requests that feel off-model, that don't fit the roadmap, that seem adjacent to your offering,
that you keep hearing and keep deferring, are often customers showing you who they actually
are and what they actually need.
Who are your best customers and how are they actually using the product?
Not your average customers, your best ones.
What are they doing differently?
If their usage pattern doesn't match your ICP assumptions, that's not an anomaly, that's
a correction.
How to fix it?
Start where this series always starts.
Actual customers.
Talk to the ones who stayed, but don't assume staying is always an active endorsement.
Some customers stay because switching costs are high.
Some stay out of inertia.
Some haven't had a reason to leave yet.
The ones worth understanding deeply are the ones who stayed actively, who chose to renew,
who referred someone, who went out of their way to engage.
They chose you when other options existed.
Why this?
Why still?
What's actually working for them that keeps them here?
Their answers will tell you more about your real ICP than any assumption you made before
you launched.
Then talk to the ones who left.
Not necessarily to win them back.
To understand what they were trying to do that the product couldn't support.
Find out where they went instead.
Ask what that product does differently and ask the harder question.
What about that product made the difference that caused them to choose it?
The exits carry information the stays can't give you.
A customer who stayed will tell you what's working.
A customer who left will tell you what wasn't there.
For product businesses, be willing to look honestly at what features are serving real customers
and what features are serving the original vision.
Those aren't always the same thing.
Maintaining features that serve nobody, costs engineering resources, creates complexity,
and obscures the parts of the product that actually matter.
For service businesses.
Test your delivery model, your scope definitions, your timelines against what clients actually
ask for.
Not what you assumed they'd want when you designed the engagement.
Then rebuild around the customer who's actually there.
Not the customer you started with.
The one reality showed you.
Everything in your product or service traces back to one question.
Is it built around what your actual customers need or around what you assumed they would
need before you had enough data to know?
Look at what's being used and what isn't.
Look at the desire paths.
The work around your customer's built because the pavement is in the wrong place.
Look at the requests that keep surfacing and keep getting deferred.
The product doesn't lie.
It tells you what assumptions it was built on.
I'm Jason Hager, and this is Marketing Notes for Entrepreneurs from Grey Leaf Media.
For leaders who believe profit and purpose don't have to be at odds.
If your product or service feels misaligned with what your actual customers need, brand
therapy helps you find where the gap is and what to do about it.
greyleafmedia.com/diagnostic
If you want to start on your own first, the ICP Toolkit is a free 15-page download that
walks you through exactly who's actually buying from you and how to build your messaging
around them.
greyleafmedia.com/findyouricp
Those are hyphen separated, so it's greyleafmedia.com/find-your-icp
And if your website isn't reflecting what you actually do for the customers who are really
buying, we build and maintain WordPress sites designed to communicate clearly to the right
people – greyleafmedia.com/services
Next time, how a wrong ICP breaks your pricing?
Why you're either leaving money on the table or pricing yourself out of the market you're
actually in?
Until then, look at what your customers are actually using.
Your customers are already telling you what the product actually is.
Most companies are still defending what they thought it would become.
[BLANK_AUDIO]
