[Hook]

Garbage in—
Vision out.
Turn the missing numbers
Into something we can shout.

[Verse]

Ninety-seven percent complete,
The progress bar looks clean.
No one asks which seven teams
Disagree on what it means.

Sales uploads a monthly file,
Finance sends the year.
Support removes the angry words
So sentiment looks clear.

Victor writes "predictive growth"
Across the sprint review.
Aisha says, "The labels drift."
Victor shrugs: "That's Data Science."

[Chorus]

Garbage in, vision out,
Polish every open doubt.
Garbage in, vision out,
Give the board a line to quote.
When the inputs fall apart,
Put a logo on the chart.

[Verse]

One segment means "enterprise,"
Another means "has paid."
A churn event is anything
The previous analyst made.

We train on stale November,
Then forecast through July.
A warning turns from amber
When the threshold moves up high.

The sprint board says "Model ready."
The model has no name.
The acceptance test reads only:
"Results support the business case."

[Chorus]

Garbage in, vision out,
Polish every open doubt.
Garbage in, vision out,
Give the board a line to quote.
When the inputs fall apart,
Put a logo on the chart.

[Heavy Breakdown]

Bad labels.
Good branding.

Red warning.
Green guide.

[Bridge]

Aisha cuts the power down.
"No claim without validation."
Victor leans in: "Can we call it
An optimisation instead?"

[Build-up]

Garbage in—

The sequencer replies.

Vision out—

The fabrication motif climbs.

[Final Chorus]

Garbage in, vision out,
Now the quiet error shouts.
Every shortcut leaves a trace,
Every guess acquires a face.
If the inputs fall apart,
The failure starts before the chart.

[Outro]

The progress bar reaches one hundred.

Nothing underneath has changed.