Yordan Ivanov

The engineers who win aren’t the best coders.


They’re the best translators.
The ones who show up with business language, not just logic.

If you can make a VP nod in the first 30 seconds,
you’re halfway to budget.

Technical skill gets you trust.
Business skill gets you power.

No one cares how it works until it breaks revenue.

Infra wins don’t resonate
until they hit numbers.

So map everything upstream:

» Faster pipeline? → On-time pricing update
» Clean data? → Fewer support tickets
» New table? → Exec dashboard without excuses

Translate tech into revenue, risk, or reach.
Then they’ll care.

Quiet execution won’t get you a raise.
Strategic friction will.

You’re not annoying when you flag issues early.
You’re indispensable.

So speak up.
Even if it’s clunky.
Even if it’s loud.

Especially if it costs the business.

Stop waiting for someone to notice.
Make the business feel it.

Every data team knows this pain:


→ Clean delivery
→ Zero pushback
→ No visibility

That’s the cost of being “low friction.”

Truth is, if your update doesn’t force a decision or a feeling…
It vanishes.

You can’t force insight with filters.
You earn it with context.

Next time you present a metric, try this:
“Here’s the trend. Here’s when it broke. Here’s what changed.”
Simple pattern recognition > complex slicing.

That’s how you shift from analyst…
To advisor.

You’re shipping. Supporting. Unblocking.
But not advancing.

That’s not because you’re slow.
It’s because the game changed, and no one told you.

This next chapter isn’t about working harder.
It’s about learning how to be seen.

Nobody asked how fast it ran.
They asked: “Why are leads missing?”

We obsess over performance metrics.
But speed doesn’t get buy-in.
Clarity does.

If your update sounds like a README file, don’t be surprised when no one reads it.

Most data engineers wait for chaos, then shine as heroes.
But heroes burn out.

Here’s the upgrade:

1/ Post a one-liner in planning docs before anyone pings you.
2/ Flag risks early.
3/ Suggest tradeoffs in business language.

Now people see you as strategic.

I once burned 40 hours optimizing Spark jobs.
Got a thank-you emoji.

- Scaled our DAGs.
- Reduced failures.
- Saved 6 hours of compute a week.

Slack said: “Thanks!”

Business impact with no business narrative = ghost effort.
If no one understands what changed nothing changed.

One engineer I coach got 2 headcount from a Slack update.
Because it made a non-technical exec look smart in a forward.

Your infra update isn’t landing?
It’s not because they don’t care.
It’s because they don’t get it.

Try this:

1/ What’s broken (1 line)
2/ What it costs (revenue, trust, hours)
3/ What you need (headcount, time, buy-in)

The business doesn’t understand our data.
Maybe.

Or maybe you built a flawless pipeline
for a problem no one cares about.

It scales beautifully.
But nobody uses it.

This is the tragedy of data engineering:

You can build something brilliant
that absolutely no one wants.

The data industry promotes builders.
Not leaders.

You nailed pipelines.
So they gave you politics.

Now you're managing people, budgets, and roadmaps.
With zero training.

You’re not unqualified.
You’re just undertrained.

Most data engineers aren’t underfunded.
They’re invisible.

You’re solving problems no one asked for.
You’re building infra no one believes in.

No one buys pipelines.
They buy certainty.

Leadership isn’t code.
It’s conviction.

If your execs don’t 'get' data, that’s your problem.

No one’s going to fund a pipeline they don’t understand or back a strategy they don’t believe in.

Writing code is not enough.
Translate complexity into certainty.

Stop proving you're smart.
Start proving you're essential.

Want all the details? Check the post here
links.datagibberish.com/g7E1

Most data work isn’t hard because of data.
It’s hard because of people.

So I’m doing something new:
A live group coaching call for Data Gibberish Pro.
Bring your messiest stakeholder drama.
I’ll bring frameworks.

🗓️ July 24.
Details in the post.

Wild how many “senior” data folks
can’t explain what they do.

Meanwhile inside Data Gibberish,
people are learning to turn noise into narrative
and getting paid for it.

This is where future data leaders are made.
You in or what?

I crashed prod last week.
No PR. No heads-up. Just ego.

Wrote a post-mortem.
Shared every detail: what broke, why, and how to fix it.

Got praised, not blamed.

Because real engineering culture = transparency > blame.
Failures taught out loud.

He left because of me.

At 26, I was a high-performing engineer and a terrible mentor.

I chased output. Ignored growth, safety, and learning.

Most new leaders think excellence is enough.

It’s not.

Excellence without empathy is a leadership bug.

I learned the hard way.

Need a guide on writing good BLUF messages? Check this out:
links.datagibberish.com/fuH7

15 slides? Too long. Didn’t read.

Execs want # by line 3.
No Bottom Line Up Front = No buy-in.

Use this:

• Problem
• $ cost
• Fix + timeline

Translate tech.
Cut fluff.
Get funded.

VP: “Make the cloud-cost dashboard prettier.”
Translation: show me €.

Slide rewrite:
• Gains: Finance & Eng
• When: next sprint
• Save: 6-figures/yr

10-min review → budget granted.

Lesson: cash beats color.

Your pipeline isn’t slow.
𝗜𝘁’𝘀 𝗹𝗲𝗮𝗸𝗶𝗻𝗴 $𝟰𝟳𝗸/𝗺𝗼𝗻𝘁𝗵.

Ask these Monday:
1️⃣ Who feels the pain first?
2️⃣ $ per min of latency?
3️⃣ Which team ships sooner if fixed?

Most data engineers aren’t taught to thrive.
Only to survive.

I’m changing that.

My newsletter teaches what no one else does:
Influence. Ownership. Career leverage.

Loving your craft isn’t enough.
You need to get seen.

Let's make this happen.
Together.

Why stakeholders don’t care about your data engineering work (yet)
links.datagibberish.com/fnLz

I rebuilt ingestion.
They said: “Launch is 2 days early.”
Same result. Different language.

They got promoted. I didn’t.

Hard work doesn’t speak. Humans do.

Invisible work doesn’t win.
Visible alignment does.

Nobody promotes a ghost.

I vibecoded a full app in a few hours.

Electron + Cursor + chaos.

→ Wrote specs in a file
→ AI acted as PM
→ Coded task-by-task

𝗣𝗿𝗼 𝘁𝗶𝗽: test each step, commit when clean.

You don’t need Jira.
You need curiosity + fast feedback loops.

Build selfishly.

𝗬𝗼𝘂’𝗿𝗲 𝗻𝗼𝘁 𝘀𝘁𝘂𝗰𝗸 𝗯𝗲𝗰𝗮𝘂𝘀𝗲 𝗼𝗳 𝗦𝗤𝗟.
You’re stuck because no one taught you influence.

I’m building the Profitable Data Engineer Framework in public
Influence. Communication. Scope. Translation.

Search Data Gibberish if you’re done being overlooked.