Capstone: Mock Loop

The screen is not a conversation. It is a filter and the two filters are your numbers and your reason for being in the room. Most candidates lose the loop here by being vague about attainment and mushy about why Cursor.

Run this section as a live drill, not a read. Set a timer, speak out loud and record it. The Cursor screen is fast and the hiring manager is a working seller who has heard every polished answer, so the bar is specificity, not smoothness.

The four-block screen rehearsalEach block is timed; over time = a cut

1. 1Why Cursor, in under 90 seconds. Name the specific thing about selling an AI-native code editor that pulls you. Tie it to a real moment, not a mission statement you read on the careers page. Land on: I want to sell to engineers and I want to sell a product devs already love.
2. 2Your numbers, in under 60 seconds. Lead with the headline: quota, attainment, ranking and how much was self-sourced. Give the denominator. "180% of a $1.1M number, 3 of 11 reps over plan, ~60% pipeline self-generated" beats "I crushed quota."
3. 3One complex deal, end-to-end, in under 4 minutes. Walk it through your qualification framework so the structure is visible. Hit metrics, economic buyer, decision process, champion and the competitor you beat. Make the stakes and the obstacle concrete.
4. 4The deal you lost. Pick a real loss, own your specific mistake and state what you changed afterward. The mistake should be diagnosable, not "the budget got cut."

Score yourself on three axes

Discovery is the stage most AEs think they're good at and most fail in the loop, because they pitch when they should question and they leave without an economic buyer or a next step.

Build the persona first, then run the clock. The interviewer in the real loop will play a technical buyer with a genuine workflow pain and at least one trap, usually security. Your job is to diagnose, quantify and exit with a committed next step, all while talking less than the buyer.

The persona to run againstMake it specific enough to surprise you

Run the 15 minutes in five moves

Grade the call against this rubric

A feature tour loses; a value-led pitch wins. The loop will test whether you can connect what Cursor does to what the buyer's business gets and then whether you hold your nerve when they push.

Build a five-minute pitch around two or three capabilities tied to outcomes you heard in discovery, then have someone fire objections at you live. The composure under "gotcha" pushback is half the grade.

Structure the five-minute pitch

1. 1Anchor on their number (30s). Open with the pain you quantified: "You said onboarding is three weeks and review is the bottleneck."
2. 2Show, tie, repeat (3 min). Demo one capability, then say the outcome out loud. Tab and codebase-aware completion against onboarding speed; Agent against the legacy-monorepo slog; review/ask-codebase against the backlog.
3. 3Quantify the case (1 min). Translate to seats, time saved and throughput, using their own numbers, not a generic stat.
4. 4Close to a next step (30s). Propose a scoped trial or POC with success criteria the champion can defend internally.

The three objections to drillPractice all three until the moves are reflex

Score the round

Cursor is known to run a multi-day in-office work trial for finalists. The GTM analog is a deliverable: a one-page account plan, a defensible ROIReturn on Investment. The value gained versus what it cost, the language an economic buyer funds deals in. model and a deal sequence you can defend live.

Treat clarity and real thinking as the test, not slide polish. A clean one-pager that shows how you'd actually attack a real account beats a thick deck of generic frameworks. Build it on a specific mid-market target so the choices are concrete.

Part one: the one-page account plan

Stakeholder map - name the committee, not just titles

Part two: a defensible ROI modelGrounded in assumed discovery numbers, stated as assumptions

Make every input an explicit, conservative assumption a CFO could challenge. The credibility is in showing your math and using their numbers, not in the size of the output.

Part three: a 30-60-90 deal sequence

1. 1Days 0-30 - convert and multi-thread. Turn the platform lead into an armed champion, run a scoped POC with success criteria and open the security thread with AppSec before they ask.
2. 2Days 31-60 - prove and expand the committee. Deliver POC results against the board mandate, get the VP Eng into a co-built business case and complete the security review with the one-pager and DPAData Processing Agreement. A contract spelling out how a vendor is allowed to handle your data. in hand.
3. 3Days 61-90 - package and close. Bring the ROIReturn on Investment. The value gained versus what it cost, the language an economic buyer funds deals in. and a phased rollout to the CTO sponsor, align Procurement and Legal on the paper and time the proposal against the Copilot renewal five months out.

One honest scorecard, scored before you walk in, is worth more than another reread of the prep. The point is to find your weakest stage and drill it, not to feel ready.

Rate each dimension 1-5 against the bar below. Be ruthless: score what you can demonstrate cold today, not what you'd do on a good day.

Find and drill your weakest stage

The readiness checklistAll four must be true before you go

• You can demo Cursor and answer the top-10 likely objections with no notes.
• You have 6-8 polished STAR stories, each mapped to a Cursor value: passion, high agency, velocity, curiosity, customer-obsession, coachability.
• Your numbers come out identically every time, denominators included.
• Every rubric dimension above is a self-scored 4 or higher.