Comparison

Podcast 2 — In conversation

Speaker A: I have been asked to write a comparison of two project management methodologies for a board presentation. I am finding it harder to structure than I expected.

Speaker B: Comparisons are structurally demanding because you have to choose between going through each item fully before moving to the next — the block method — or addressing each criterion in turn for both items simultaneously — the point-by-point method. For a board presentation, point-by-point tends to work better because the audience can immediately see the differences per criterion without holding two complete profiles in their head at once.

Speaker A: That is very helpful. I have been using the block method and I think that is why it feels clunky.

Speaker B: The block method works well in writing where readers can page back. In a live presentation, point-by-point is clearer because it makes the comparison visible in real time.

Speaker A: What about the phrase by comparison? I keep writing it but I am not entirely sure I am using it correctly.

Speaker B: By comparison introduces a relative assessment — it signals that what follows is being measured against something already established. The new system is faster — and by comparison with the legacy platform, the improvement is dramatic. You need the reference point to have been introduced before you use by comparison, otherwise the phrase floats without an anchor.

Speaker A: So by comparison requires a prior point of reference.

Speaker B: Exactly. It is inherently backward-looking — it reaches back to something already stated. In comparison with is slightly different: it can introduce the reference point directly. In comparison with last quarter, this quarter's results are encouraging — the reference point is in the same clause.

Speaker A: What about there is no comparison? I want to say that one methodology is clearly superior.

Speaker B: There is no comparison is one of the most emphatic ways to express superiority — it dismisses the very possibility of meaningful competition. It is strong and slightly informal, so in a board presentation you might prefer there is a clear advantage or the difference is substantial. But in conversation, there is no comparison lands with real force.

Speaker A: I shall save there is no comparison for the informal debrief and use clear advantage in the report.

Podcast 3 — Prompt engineering

Speaker B: Comparison is a powerful noun for structuring analytical and reporting features in prompts. It names the output, not just the action — a comparison report, a comparison view, a comparison score. Let me show you how to use it precisely.

Speaker A: I tend to phrase things as show the difference between — is comparison a better choice in prompts?

Speaker B: Comparison is more specific because it names the output artifact — a comparison is a structured product, not just a visual arrangement. First — a financial reporting tool:

Speaker B: PROMPT:

Speaker B: Comparison as the report type immediately shapes the data model — budget, actual, variance — and the output format. The AI knows this is an analytical artifact, not just a display.

Speaker A: Budget versus actual comparison — that is a standard accounting view. The word comparison drives the whole structure.

Speaker B: Database query:

Speaker B: PROMPT:

Speaker A: Year-on-year comparison as a named query type — the label growth, decline, stable is a real analytical classification. The word comparison frames the intent clearly.

Speaker B: Now a UI feature for an HR tool:

Speaker B: PROMPT:

Speaker A: Comparison panel as a named UI component — that is a real product term. The comparison summary row is a computed output driven by the comparison.

Speaker B: One more — full application:

Speaker B: PROMPT:

Speaker A: Historical comparison snapshots — that is a time-series dimension built directly into the product model. The word comparison creates a named entity that can be stored, retrieved, and tracked.

Speaker B: Use comparison in prompts whenever you want the AI to build a structured analytical artifact — a named output with columns, scores, labels, and recommendations. It consistently produces more complete, production-ready analytical features than vaguer instructions like show differences.