Wall Street PromptBrand
All briefs
Wall Street Prompt · Brand

Claude In Excel Is Not The Point

Carousel — cover + one idea per slide + caption. This is a post, not a video script.

Carousel
Instagram carousel + LinkedIn documentInstagram carousel8 slides
References & validationExact post missing

Reference creator resolved, but the exact source post was not captured. Capture the specific post (not just the profile) before treating as validated.

Reference creator: Nicolas Boucher

AI Finance Club · LinkedIn

Slides

One idea per slide

Cover
CLAUDE IN EXCEL
IS NOT THE POINT

The feature lives in the cell.
The edge lives in the review.

Model review. Assumption challenge.
Source tie-out.

AI builds. You judge.
(swipe)
Slide 1
CLAUDE IN EXCEL
IS NOT THE POINT

The feature lives in the cell.
The edge lives in the review.

3 shifts that actually matter.
AI builds. You judge.
(swipe)
Slide 2
Everyone is talking about the same thing:
"AI can build the model now."

That is the small part.

A model built faster is not a model built right.

The harder question:
who is checking the work?
Slide 3
THE REFRAME

The old job was speed.
Build the model faster.

The new job is review.
Find what breaks the model.

-> Building is now cheap
-> Checking is the scarce skill
-> Judgment is the part that does not scale
Slide 4
SHIFT 1: MODEL REVIEW

Stop asking AI to build it.
Ask AI to break it.

-> Where does a formula grab the wrong row?
-> Where is a hardcoded number buried in a formula?
-> Where does a sign flip or a unit change?
-> Where does a total stop tying?

You read the flags. You decide what is real.
Slide 5
SHIFT 2: ASSUMPTION CHALLENGE

Every input has to be defended.

-> Why this growth rate?
-> Why this margin path?
-> Why this discount rate?
-> Why does this hold through a downturn?

"It was in the template" is not a reason.
AI lists the inputs. You defend each one.
Slide 6
SHIFT 3: SOURCE TIE-OUT

Every number traces to a source.

-> Revenue ties to the filing
-> Guidance ties to the transcript
-> Comps tie to a named, dated set
-> Anything else is a stated assumption, labeled as one

A page and a date, or it leaves the model.
Slide 7
WHAT AI IS NOT DOING

-> Not deciding the company is cheap
-> Not telling you what it is worth
-> Not signing off on your assumptions
-> Not owning the call

It fills the cells faster.
The thesis is still yours.
Slide 8
THE POINT

The cell is the feature.
The review is the edge.

-> Model review: ask it to break the model
-> Assumptions: defend every input
-> Tie-out: source and date, or cut it

AI builds. You judge.
Save this. Run it on your next model.

Educational only. Not advice.

Caption

Paste under the carousel

Claude in Excel is a real upgrade. It is also not the point.

The feature lives in a cell. The edge lives in what you do after the cell is filled. Building a model used to be the slow, skilled part. Now building is cheap and review is the scarce skill.

Three shifts, AI builds and you judge:

1. Model review: stop asking AI to build the model, ask it to break it. Wrong row references, hardcoded numbers inside formulas, sign flips, totals that stop tying. 2. Assumption challenge: every input gets defended. Why this growth rate, this margin, this discount rate. "It was in the template" is not a reason. 3. Source tie-out: every number traces to a filing, a transcript, or a labeled assumption. A page and a date, or it comes out of the model.

Notice what the AI is not doing: it is not telling you the company is cheap, what it is worth, or whether to act. That stays with you.

Save this one and run it on your next model.

Educational content only. Not investment advice, and not a recommendation to buy, sell, or hold any security. Wall Street Prompt. Always verify against the primary source.

Design notes

Near-black slides (#0B0B0C), one accent: WSP green (#19C37D). Cover uses heavy condensed all-caps headline (Anton / Druk), green on "IS NOT THE POINT". Body slides: a small green section label at top ("SHIFT 1: MODEL REVIEW"), a one-line concede-then-sharpen statement, then a dense arrow-list (-> bullets) as the teaching unit. One idea per slide, lecture-slide density, not minimalist tip-cards. Arrows are typed -> in green, body text in white grotesk (Inter / Söhne) at readable size with generous leading. Consistent left-aligned grid, fixed top margin for the label, page number small bottom-right, WSP handle small bottom-left. Final slide repeats the three-shift arrow-list as a recap plus the "AI builds. You judge." mantra and the disclaimer in smaller type. No icons, no stock images, no gradients. Built to be saved and swiped back through.

CTA

Save this and run it on your next model. Comment "TIE-OUT" and I'll send the model-review checklist.

Wall Street Prompt
Wall Street Prompt — internal