Fancy UI isn’t UX. Real UX leaves evidence.

[Photo content: abstract heatmap or blurred analytics dashboard]

Most teams say they “care about UX.”
Very few can prove it works.

Design reviews are full of opinions.
Stakeholders vote.
Designers defend.
Someone says, “Users will like this.”

And then the product ships — without a single baseline, index, or success metric.

If you can’t measure how your design changed user behavior, you didn’t design UX.
You designed a guess.

UX Is Not an Opinion. It’s a Result.

UX does not live in Figma.
UX does not live in Dribbble shots.
UX does not live in presentations.

UX exists after release, when users interact with your product under pressure, distraction, and zero patience.

If nothing changed in how users behave, decide, or complete tasks — UX did not improve.

“Looks better” is not a UX outcome.
“Feels cleaner” is not a UX metric.

Real UX answers only one question:

Did user behavior change in a meaningful way?

Why Most Teams Can’t Measure UX

Not because they don’t have tools.
Because they don’t have clarity.

Common failures:

• No baseline before redesign

• No definition of “success”

• Metrics chosen after launch

• Engagement confused with understanding

• Tools used without interpretation

Teams collect data, but don’t ask the right questions.

If everything is a metric, nothing is insight.

The UX Indexes That Actually Matter

Vanity metrics don’t measure UX.
Indexes do.

These are the signals that reflect real user experience:

• Task Success Rate
  Can users complete what they came to do?

• Time on Task
  Not speed — clarity. Confusion slows people down.

• Error Rate
  Misclicks, form failures, dead ends.

• Drop-off Points
  Where intent disappears.

• Rage Clicks
  Frustration disguised as interaction.

• Scroll Depth vs Action
  Scrolling means nothing if users don’t act.

• Return Behavior
  Do users come back without being pushed?

UX is found in patterns, not single numbers.

Tools Don’t Do UX. People Do.

Hotjar.
Microsoft Clarity.
FullStory.
GA4.
Maze.
UsabilityHub.

All useful. None magical.

A heatmap without context is just modern art.
Clicks don’t equal intent.
Sessions don’t equal success.

Tools show what happened.
UX thinking explains why it happened — and what to change.

Used wrong, research tools create confidence without truth.
Used right, they expose uncomfortable reality.

Sources & References (Real UX Case Studies)

1. Baymard Institute — Cart Abandonment Research
In-depth e-commerce UX data showing how specific UX issues drive shopping cart abandonment and affect revenue.
🔗 https://baymard.com/lists/cart-abandonment-rate

2. Google Web Vitals — UX Performance Metrics
Official documentation and research on Core Web Vitals and how performance affects user behavior and business outcomes.
🔗 https://web.dev/vitals/
🔗 https://web.dev/why-speed-matters/

3. GOV.UK Service Manual — Measured UX Improvements
UK government’s design system and service standards with evidence of improved task completion and reduced support costs.
🔗 https://www.gov.uk/service-manual

4. Microsoft Clarity — Behavior Signals (Rage Clicks)
Documentation and insights on how rage clicks and other behavior signals can reveal UX frustration points.
🔗 https://learn.microsoft.com/en-us/clarity/

5. Nielsen Norman Group — How Users Read on the Web
Research on scanning patterns, visual hierarchy, content comprehension, and user reading behavior.
🔗 https://www.nngroup.com/articles/how-users-read-on-the-web/

6. NN/g — UX Metrics Best Practices
Guidance on meaningful UX metrics and how to measure task success, error rates, and behavioral outcomes.
🔗 https://www.nngroup.com/articles/ux-metrics/

7. Baymard Institute — Form Usability Findings
Detailed findings on form design issues and how they affect completion rates — useful for real UX measurement.
🔗 https://baymard.com/research/form-usability

8. Google — Search and UX Research (User Behavior)
Broad UX research and best practice insights from Google’s UX teams and behavioral studies.
🔗 https://research.google/teams/brain/ux/

Not assist.
Not change workflows.
Replace.

And yes — something is being replaced.
Just not what most people think.

Companies Won’t Fire Designers — They’ll Shrink Teams

Let’s get real.

In most companies, design teams were bloated long before AI showed up.

Where there used to be:

• 8–10 designers doing repetitive UI work
  There will likely be:

• 3–5 designers doing higher-level thinking

AI doesn’t eliminate design.
It eliminates redundant execution.

Layouts, variations, quick visuals, internal drafts — that work gets automated.
Decision-making does not.

This pattern already exists.

Multiple industry studies (McKinsey, BCG, Deloitte) show the same trend:

• Fewer people

• Higher responsibility

• Stronger emphasis on judgment, brand thinking, and systems

AI compresses teams.
It doesn’t remove the need for taste.

Execution Is Cheap Now. Taste Isn’t.

For years, design was sold as:

“I can make it look good.”

That era is over.

AI can already:

• Generate layouts

• Suggest typography

• Produce visual styles

• Mimic trends instantly

What it cannot do:

• Decide what matters

• Understand brand values deeply

• Say “this doesn’t belong here”

• Protect clarity when everything wants attention

That gap is growing — not shrinking.

Agencies Won’t Disappear — Bad Ones Will

There’s a myth that AI will kill agencies.

Reality is harsher:
It will kill agencies without point of view.

Agencies that survive will be the ones who:

• Understand brand values beyond visuals

• Know what a brand should never do

• Design systems, not just screens

• Guide decisions, not just deliver assets

Clients don’t hire agencies for pixels.
They hire them for direction.

AI can generate options.
Brands still need someone to choose.

Confidence Is Becoming a Business Skill

In an AI-heavy workflow, the most valuable designer is not the fastest.

It’s the one who can say:

“Stop. This is enough.”

Confidence now means:

• Fewer options

• Stronger decisions

• Clear boundaries

Designers who hide behind complexity, endless iterations, or “just one more version” will struggle.

AI loves infinite versions.
Good design hates indecision.

The Price of Real Design Will Split

Here’s the part no one talks about.

Design pricing won’t just go down.
It will break in two.

1. Cheap, fast, automated design

   • Templates

   • AI-generated visuals

   • Commodity UI

2. Human-led, taste-driven design

   • Brand systems

   • Strategic UX

   • High-stakes decisions

The middle disappears.

And yes — the price of real human design can increase, because:

• Fewer designers operate at that level

• Stakes are higher

• Mistakes are more expensive

When design influences trust, conversion, and brand survival — automation alone isn’t enough.

This Isn’t About Dribbble. It’s About Judgment.

This isn’t about tools.
It’s not about Figma, Midjourney, or the next model.

It’s about:

• Knowing when to remove

• Knowing what aligns with the brand

• Knowing what feels wrong even if it “looks right”

AI doesn’t have taste.
AI doesn’t have responsibility.
AI doesn’t carry consequences.

Humans do.

Who Will Actually Lose

Let’s be honest.

AI will replace designers who:

• Only execute instructions

• Rely on trends instead of thinking

• Confuse decoration with value

• Avoid responsibility for decisions

That was fragile work even before AI.

Who Will Win

Designers who:

• Understand brands deeply

• Design for focus, not attention

• Make fewer, stronger decisions

• Trust their judgment

AI becomes their accelerator — not their threat.

Final Thought

AI is not the end of design.

It’s the end of:

• Indecision

• Overstaffed execution

• Design without taste

The future isn’t tool-driven.
It’s judgment-driven.

If you have taste and confidence,
AI doesn’t replace you.

It clears the room.

Design Centered, Not Overloaded

The entire experience is design-centered and minimal by intent.
We actively avoid cognitive overload by reducing unnecessary visual noise, excessive text, and complex interaction patterns. Every element on the page has a clear purpose.

• No cluttered layouts

• No competing visual hierarchies

• No decorative distractions disguised as “creativity”

Typography Built for Readability

Typography on verybad.agency is optimized for readability first:

• Clear, modern sans-serif typography

• Generous spacing between lines and sections

• Predictable text structure and hierarchy

• Short, scannable text blocks instead of dense paragraphs

The goal is simple: text should be read, not decoded.

High Contrast for Visual Comfort

We use a black background with white text deliberately — not as a trend, but as a usability decision.

• Strong contrast improves letter recognition

• Reduces visual fatigue

• Helps users maintain focus while reading

This contrast supports dyslexic users and benefits everyone, especially during long or repeated visits.

Motion Without Cognitive Load

Animation is used with restraint and intention.

Our motion design:

• Focuses on showing services, not distracting from content

• Avoids rapid, looping, or chaotic motion

• Keeps transitions predictable and purposeful

To maintain orientation and reduce cognitive strain:

• Section headers remain consistent across animated service blocks

• Users always know where they are and what they are viewing

Motion enhances understanding — it never competes with it.

Accessibility Is Functional, Not Cosmetic

We do not rely on “accessibility overlays” or surface-level fixes.
Instead, accessibility is embedded into:

• Layout logic

• Content structure

• Visual hierarchy

• Interaction flow

This approach aligns with WCAG principles and modern UX laws focused on clarity, consistency, and cognitive load reduction.

Designed for Real Humans

Dyslexic-centered design improves usability for:

• Users with dyslexia

• Users with ADHD or cognitive fatigue

• Non-native language readers

• Anyone navigating content quickly or under pressure

Inclusive design is not niche — it’s better design.

We hired a chatbot.
Not because we’re lazy. Because we’re obsessed with control and chaos at the same time.

When we launched Very Bad Agency, we wanted branding that felt like a drug. Addictive. Dangerous. Underground.
But building that voice? Hell. It’s hours of copywriting, brainstorming, and explaining “why we can’t use another generic tone.”

Then came ChatGPT. Everyone was using it like a homework cheat code. We didn’t want that.
We wanted a brand manager that doesn’t complain, doesn’t sleep, and doesn’t ask for “alignment meetings.”

Step One: Realization

AI doesn’t create your brand. You do.
AI doesn’t give you taste. It only reflects yours.
If you treat ChatGPT like a creative god, you’ll get soulless garbage.
If you treat it like an obedient intern with an attitude problem, you win.

So we gave it direction. Hard, strict, non-negotiable direction.

Our Path of Prompts

We didn’t start with “Write me Instagram captions.”
We started with a manifesto:

“You are not polite. You are not a corporate voice. You are raw, rebellious, honest. You speak like an underground cult. Your tone is addictive like drugs. Your purpose is to make design feel illegal.”

We taught it:

• The mood → edgy, bold, zero fluff.

• The language → mix of sarcasm, brutal honesty, and shock value.

• The rules → no fake positivity, no “we’re thrilled to announce,” no marketing clichés.

Every time it drifted back to “Hi there! 😊,” we slapped it with more context.
Brand AI is like a dog—train it or it pees everywhere.

What AI Does for Us

• Keeps the tone consistent across socials, blogs, and proposals.

• Speeds up ideation for campaigns like Daily Dose and Bad Examples™.

• Brainstorms without judgment (unlike that one intern who thinks he’s Don Draper).

But here’s the truth:
AI isn’t “creative.”
Creativity = taste + decision-making.
ChatGPT just amplifies what you feed it.

Our Attitude Towards AI

AI isn’t scary.
Bad taste is.

Most agencies fear AI because they think it replaces jobs.
We fear agencies that think AI makes creativity effortless.
No. It makes execution fast, but direction is still on you.

Our job? To stay on the correct branding path—even when AI tries to lure us into “safe” territory.
(And it will. AI loves mediocrity if you let it.)

The Rule We Live By

“AI should sound like us. Not like AI.”

If the copy feels like it came from a sterile content mill, we kill it.
Every post still passes the Very Bad Test:

• Does it feel underground?

• Does it hit like a punch, not a pillow?

• Does it make someone uncomfortable in the best way?

If yes, publish.
If no, we throw it back into the digital void.

What We Learned

• Prompts are everything. The first thing you write defines the brand.

• Consistency beats creativity. AI helps maintain that over dozens of posts.

• Direction is king. Without attitude, AI will default to LinkedIn corporate hell.

So, did AI replace our brand manager?
No.
We turned it into one—by giving it taste, rules, and zero room for compromise.

Because a brand without taste is just another Instagram account begging for likes.
And we’re not here to beg.

Follow our chaos:
Instagram
LinkedIn

Back in the lab, it was just two guys with problems:

• Tengo Khunashvili — UX/UI designer who obsesses over flows and taste.

• Shota Datashvili — Motion designer who sees in frames, not seconds.

We weren’t chasing clients. We weren’t building pitch decks.
We were building our portfolio. But even that felt too basic.

So we decided to go harder.
A full-on visual identity, from scratch.
Logo. Name. Website. Motion. Vibes.
A whole damn brand, just for us.

But we didn’t wanted another “clean aesthetic” that dies on launch day.

We want something that slaps. Something brutally honest.
Something that walks in the room and offends half the crowd (the right half).
And for three long months, we had… only the name:

🚫 Very Bad Agency

No visuals. No philosophy. No soul.
Just a provocative name and a shared obsession for doing things better by being worse.

And the hardest part?
We couldn’t even start the website.
Because the brand had no voice. No body.
You can’t design a homepage for a ghost.

💡 Then It Hit Us: Digital Pharmacy

Two words changed everything.

Digital. Pharmacy.
That was the key.
The medicine for boring brands.
The prescription for dead design.

Suddenly, it all clicked:

• Pills for services

• Doses for content

• Dealers for strategists

• Injections for high-impact branding

• And one glorious side effect: Addiction to taste

We built the site in under 48 hours, with Spline 3D + Webflow.
Raw, animated, unapologetic.

Then came Instagram.
But no basic case studies or filtered quotes.
We dropped Daily Doses — savage truth capsules for digital creatives who needed a slap, not a hug.

Examples?

• “Rest isn’t lazy. It’s strategy.”

• “You don’t need better tools. You need better taste.”

• “No one’s coming to save you. Good. Build without permission.”

When others optimized for clicks, I optimized for meaning.
we made Awakening Reels — audio-visual shocks for the procrastinating brain.
They weren’t optimized.
They weren’t polished.
They were real.
That’s why they worked.

🚧 What Now?

We’re not done.
We’re not even stable yet.
We're actively switching strategy, because building something raw doesn’t mean being reckless.

This article?
Yeah, it’s part of our SEO plan.
We’re saying it out loud because that’s what honesty looks like.

But unlike most agencies who pretend to be something they’re not,
we built this whole thing in public — through conversations, struggles, and a stubborn refusal to settle.

We’re still underground. Still dangerous.
Still learning how to weaponize taste in a market that rewards blandness.

And if you’re reading this and still thinking,

“Wait, is this a joke or a real agency?”

Here’s your answer:

It’s both.
And we’re Very Bad at pretending otherwise.

Want to feel something again?
→ verybad.agency

→ Instagram