A topic is not a structure
"AI in operations", "what I learned from biographies", "my business journey", and "delegation" are topics. A topic tells the writer what material is nearby. It does not tell the reader why one section follows another.
Aristotle's most useful structural idea in your reference document is that unity comes from one complete action, not from one person or one subject. For argumentative writing, the equivalent is one governing question or claim.
A governing action contains movement
Topic:
"Building an internal ERP."
Governing action:
"We tried to automate payroll before fixing the attendance truth chain; the failure forced us to redesign the system around evidence rather than data entry."
Topic:
"Hiring a sales leader."
Governing action:
"I delegated a process before I knew how to measure it, then discovered I could not distinguish poor execution from my own bad expectations."
The removal test finds sections that do no work
Ask this of every major section:
If I remove this, does the reader reach the same conclusion by the same route?
If yes, the section may be one of these:
- decoration
- duplicated evidence
- background arriving too early
- a second essay
- an example that proves nothing new
Do not use the removal test mechanically. A qualification may leave the route unchanged but still change the ethical or epistemic strength of the conclusion. That makes it structural.
A long piece must compress into a short causal spine
A long essay can contain many episodes, as long as the reader can compress the whole into a short causal spine.
Write the piece in five to seven sentences. If those sentences contain three unrelated plots or conclusions, the full essay will probably sprawl.
Your Aristotle reference gives a useful rule:
Compress the spine; expand the evidence.