Every few weeks I have to turn a pile of competitor notes—screenshots, pricing pages, feature lists, and scattered observations—into a clean presentation outline for an internal meeting. The research itself is rarely the hard part. Turning that research into a logical story someone can follow in ten slides is.

I tested whether AI could handle the jump from messy notes to a usable outline. Here is how it went.

The Task and Constraints

What I needed

A presentation outline that covered:

• Three main competitors

• Key points of comparison (positioning, strengths, gaps)

• What mattered most for our current projects

• Clear section breaks that could become slides

The audience was internal. The tone needed to stay factual and calm—no hype, no exaggerated claims. All company and competitor specifics in this write-up are generalized.

Test conditions

• Tool: the AI writing assistant I currently keep as my daily driver

• Input: a working document of rough notes, bullet observations, and copied feature lists (about 1,200 words of messy material)

• Time limit: 30 minutes total for prompting, revisions, and human cleanup

• Goal: an outline I would be willing to drop into slides without restructuring the whole thing

The Workflow

First prompt: structure the material

I pasted the notes and asked for a presentation outline with:

1. A short opening frame

2. One section per competitor

3. A comparison section

4. A closing section on implications for our work

I also included the usual constraints: no invented claims, plain language, and a length that would fit roughly ten slides.

The first output gave me clean headings and a logical order. It also smoothed over two important distinctions in the notes and treated one minor feature as more significant than the source material supported.

Second pass: restore accuracy and emphasis

I pointed out the overweighted feature and the blurred distinctions, then asked the tool to re-rank the points strictly according to the notes I had supplied. The second version corrected the emphasis and produced tighter bullets.

Third pass: slide-ready language

I asked for a final version in which each main bullet could sit on a slide with minimal rewriting. This pass cleaned up the remaining soft transitions and gave me something close to usable.

What the AI Handled Well

• It imposed order on scattered notes quickly.

• Section logic (competitor-by-competitor, then comparison, then implications) appeared on the first try.

• Once corrected, the outline stayed stable across small follow-up requests.

• The jump from raw research to a narrative skeleton saved the usual twenty minutes of staring and rearranging.

What Still Required Human Work

Judgment about importance

The tool cannot know which competitor detail actually matters for our current priorities. It can only work with the weight I give it in the notes or in follow-up instructions. I still had to decide what deserved a full slide and what belonged in a single bullet.

Accuracy against source material

Even with clear instructions, the first draft slightly inflated one strength and softened one gap. Clean formatting made the changes easy to miss. A careful read against the original notes remained necessary.

Internal context the notes do not contain

Recent conversations, unspoken priorities, and political sensitivity never appear in the research document. Those filters stayed with me.

Final slide design and evidence

The outline did not include screenshots, exact pricing tables, or source links. Those still had to be added by hand.

Time Result

• Prompting and three AI passes: 11 minutes

• Human accuracy check, emphasis decisions, and light rewriting: 14 minutes

• Total: about 25 minutes

My previous average for the same kind of outline starting from raw notes: 40–45 minutes.

The savings came almost entirely from skipping the blank-page structuring phase.

The Verdict

I will keep using AI for the first pass when I need to turn competitor research into a presentation outline. It reliably produces a logical skeleton and reacts well to specific corrections.

I will not accept the first outline as final. Emphasis, accuracy, and relevance to our actual work still require a human pass. The tool is good at organizing what I give it. It is not good at knowing what matters most.

Clear limitation

This approach works when the research notes are already reasonably complete. If the notes themselves are thin or biased, the outline will simply organize the weakness more neatly. The quality of the input still sets the ceiling.

I tried it so you don’t have to waste your afternoon.