Capstone: Full Mock Loop
Run the whole gauntlet and grade yourself to the bar
How to run a realistic mock loop
After this you can After this you can set up a self-exam that mirrors the real stages and pressure.
You do not rise to the level of your prep. You fall to the level of your reps. A mock loop only works if it feels uncomfortably close to the real thing - same clock, same hostility, same no-AI rules.
The Cursor SDR loop has five live stages and each one breaks a different muscle. Reading about cold calls will not make you better at cold calls. The point of this module is to put the whole gauntlet on a calendar and run it against a partner who is willing to be unpleasant.
Interactive diagram. Tab through its regions; each focused region shows its detail in the panel below.
Step through each stage to see what it tests and how to brief your partner.
Build the loop before you run itOne setup pass, then you can rerun any stage cold
- 1Recruit a partner - or record yourself. Ideally one human who can switch hats: warm recruiter, skeptical engineering buyer, metrics-driven hiring manager, then an AE. If you are solo, run each stage to a recording and grade the playback. A real voice on the other end is worth a lot, so trade reps with another candidate if you can.
- 2Brief them on the persona. Hand your partner one paragraph per role: what to push on, what objections to throw, what a strong answer sounds like. A buyer who is randomly nice teaches you nothing.
- 3Time-box every stage to the real length. A 30-minute screen runs 30 minutes. The cold call is one live shot with no restart. The take-home gets a hard deadline and a closed tab.
- 4Sequence in the real order. Screen, then call, then take-home, then behavioral. Stamina is a skill - answering your fifth hard question of the day is different from answering your first.
- 5Capture an artifact from each stage. A call recording, the actual email you wrote, your spoken answers. You cannot grade what you did not record and memory flatters you.
- Recruiter + HM screen
- 30 min, spoken, no notes read aloud
- Live cold call
- 5–8 min, one take, partner plays a hostile engineer
- Written take-home
- 45–60 min, hard deadline, no AI assist
- AE / cross-functional
- 20 min working session reviewing your take-home
- Values / why-Cursor
- 20 min, recorded, conviction under probing
Run them in this order in one sitting once you can survive each alone.
Cursor screens hard for whether you actually use and love the product. Its CEO has said candidates who treat it 'just as a job' will not pass. So your mock has to respect a real Cursor norm: do the writing and the role-play unaided. If you lean on an AI to draft your cold email in practice, you are rehearsing a skill you cannot use in the room and you are dodging the exact thing they want to see - your own judgment and voice.
Each stage is testing the same person under a different lens. The recruiter checks motivation, the call checks composure, the take-home checks your writing, the AE checks whether they would trust your pipeline. Running them back to back surfaces the cracks that only show up when you are tired - the rushed close, the over-long answer, the email typo at minute 50.
Takeaway. A useful mock is a realistic mock: real clocks, a partner briefed to be hostile, no AI assist and a captured artifact you can grade against the bar.
Self-check
QWhy is it important to do the written take-home unaided in your mock, even though no one is watching?
Mock 1 - recruiter + hiring-manager screen
After this you can After this you can rehearse your motivation, fit and sales-IQ answers out loud and time them.
The screen is a filter, not a chat. Two filters decide it: a specific reason you want this seat and numbers that prove you understand the job. Vague loses here.
Run this as a live drill. Set a timer, speak out loud, record it. The recruiter checks motivation and genuine product affinity. The hiring manager - a working seller - checks sales IQ, self-sourcing and whether the AE path is a real goal or a throwaway line.
The questions to drill, each timedLong answers read as poor SDR instincts
- Why Cursor? (under 75 sec) - a real reason rooted in using the product, not a careers-page mission statement. Name what you do in Cursor and why automating coding pulls you.
- Why sales and why SDR now? (under 60 sec) - own the choice; do not apologize for starting at the bottom of the funnel.
- Comp and location (under 30 sec) - know the SF/onsite expectation and your number; a fumble here reads as unserious.
- Your self-sourcing story (under 90 sec) - one concrete time you built pipeline or a network from scratch without leads handed to you.
- Pipeline math (under 60 sec) - walk activity to conversations to SQLs to opps with real ratios.
- "You're behind quota - what do you do?" (under 75 sec) - diagnose with data, then act; do not just promise to work harder.
On the AE path: "I want to be an AE and SDR is the fastest way to earn it - I learn the ICP, the objections and the product by living in outbound first. I'm not looking to skip the rep work; I want to be the SDR an AE fights to get pipeline from and grow into closing from there."
Pipeline math you should be able to recite
- Activity
- ~60 dials + 40 personalized emails/day
- Conversations
- ~5 real connects from those dials
- Meetings booked
- ~2–3/week from conversations + replies
- SQLs / qualified opps
- ~6–10/month handed clean to an AE
- The lever
- More quality conversations, not just more dials
Numbers are illustrative; the point is you can reason about the ratios.
When asked about being behind quota, lead with the diagnosis before the hustle. "First I'd look at where the funnel is leaking - if connects are fine but meetings aren't booking, my opener or qualification is the problem, not my activity. Then I'd A/B one variable and check it against the data in a week." This shows the data-driven, iterate-fast instinct Cursor wants far better than "I'd grind harder."
On the playback, mark every "so basically," "kind of," "to be honest," and every run-on second answer to a question you already answered. Cut them and re-record until your "why Cursor" and your numbers come out the same crisp way each time. Under loop pressure you default to your rehearsed version, so make the rehearsed version tight.
Takeaway. Win the screen on specificity: a product-rooted "why Cursor," numbers with real ratios and an AE-path answer that respects the rep work.
Self-check
Mock 2 - the live cold call
After this you can After this you can run a full cold call to a hostile engineering buyer and survive objections.
This is the job, live, under pressure. The mock cold call is the stage where most candidates either pitch over the buyer or fold at the first "we already use Copilot." Your edge is composure and talk-time discipline.
Have your partner play a skeptical engineer who may already know Cursor better than you do. One take, no restart. Your job is a permission-based opener, real discovery, calm objection handling and a specific meeting ask - while the buyer talks roughly 80% of the time.
Run the call in five movesBrief your partner to interrupt and resist
- 1Permission-based opener (15 sec). Be honest it's a cold call, then earn 30 seconds. "I know I caught you cold - can I take 30 seconds to say why I called and you tell me if it's worth more?"
- 2Pattern interrupt + reason for the call (20 sec). Lead with a researched trigger, not a pitch. "I saw your team's hiring a bunch of backend engineers - usually that means onboarding and review load spikes."
- 3Discovery (the bulk). Ask, then shut up. Get to how the team actually writes and reviews code today and what AI tooling they've tried. Quantify the pain if you can.
- 4Objection handling (as they come). Acknowledge–Ask–Advance. Never argue with an engineer about their own stack.
- 5Specific meeting ask (15 sec). Not "some time to chat" - a named slot with a named person. "Worth 20 minutes Thursday with our AE who works with teams your size? I'll bring the two things most relevant to onboarding."
Interactive diagram. Tab through its regions; each focused region shows its detail in the panel below.
The meeting ask is the gate - everything before it exists to earn the right to ask.
The hard objections your partner must throw
- Objection
- "We already use Copilot."
- Weak response
- "Cursor is way better than Copilot."
- Acknowledge–Ask–Advance
- "Makes sense - a lot of teams start there. What's the one thing you wish it did that it doesn't? That gap is usually why folks take a look."
- Objection
- "We're not switching tools."
- Weak response
- "You really should, it's worth it."
- Acknowledge–Ask–Advance
- "Totally fair, switching is a pain. Most teams don't switch - they try it on one repo first. Curious what would even make a new editor worth the friction for you?"
- Objection
- "Just send me an email."
- Weak response
- "Sure, what's your email?"
- Acknowledge–Ask–Advance
- "Happy to - so I send something actually useful and not noise, what's the workflow that eats the most time right now? I'll point the email at that."
- Objection
- "No budget for this."
- Weak response
- "It's not that expensive."
- Acknowledge–Ask–Advance
- "Got it, budget's tight everywhere. A lot of teams start free, bottom-up - if a few engineers loved it, who'd even be in the room for a budget conversation later?"
| Objection | Weak response | Acknowledge–Ask–Advance |
|---|---|---|
| "We already use Copilot." | "Cursor is way better than Copilot." | "Makes sense - a lot of teams start there. What's the one thing you wish it did that it doesn't? That gap is usually why folks take a look." |
| "We're not switching tools." | "You really should, it's worth it." | "Totally fair, switching is a pain. Most teams don't switch - they try it on one repo first. Curious what would even make a new editor worth the friction for you?" |
| "Just send me an email." | "Sure, what's your email?" | "Happy to - so I send something actually useful and not noise, what's the workflow that eats the most time right now? I'll point the email at that." |
| "No budget for this." | "It's not that expensive." | "Got it, budget's tight everywhere. A lot of teams start free, bottom-up - if a few engineers loved it, who'd even be in the room for a budget conversation later?" |
The pattern: validate, ask a question that re-opens, then advance - never defend.
Mid-call, ask your partner for one piece of feedback and apply it on the spot - "too much talking? let me hand it back to you" - then visibly change behavior. Coachability is a stated thing Cursor wants and showing you can take a note in real time beats running a clean call that never adjusts.
The fastest way to fail the call is to win the argument. If an engineer says Copilot is fine, do not correct them - an engineer may know Cursor better than you do and arguing torches credibility. Your honest edge is curiosity about their workflow, not a feature-by-feature takedown.
Before you debrief with your partner, write down one strength and one fix from your own read. Truth-seeking is a Cursor value and an honest self-assessment of a call you just ran is exactly the muscle the values round will test later. Then compare your read to your partner's - the gap between them is your blind spot.
Takeaway. Survive the call by holding ~80% talk-time on them, running Acknowledge–Ask–Advance instead of arguing and asking for a specific meeting, not vague time.
Self-check
QThe engineer says, "We already use Copilot and it's fine." Which response best fits a skeptical-engineer cold call?
Mock 3 - the timed written take-home
After this you can After this you can produce a researched cold email plus LinkedIn message under a deadline.
The take-home is Cursor's first written impression of you and it is graded on personalization, clarity, no-hype framing and a crisp CTA. Set a 45-minute timer, close the AI tab and treat a typo as a rejection.
Pick a real engineering org, find a genuine trigger and choose a named persona. Then write one tight cold email with a single CTA and a short LinkedIn message. The skill being graded is making a complex AI dev tool simple and relevant to an engineer in your own words.
The 45-minute clock
- 1Research (15 min). Pick a target account and a real trigger - a hiring spree, a migration, a public eng-blog post, a new product launch. Choose the persona: a staff/backend engineer, an eng manager, a VP Eng.
- 2Draft the email (15 min). One trigger-based reason for reaching out, one relevant value point in their language, one CTA. Keep it under ~90 words.
- 3Draft the LinkedIn message (5 min). Shorter and warmer; reference the same trigger, lighter ask.
- 4Proofread to zero typos (5 min). Read it out loud. A single typo undermines the "clear, concise writing" the role requires.
- 5Self-grade and rewrite once (5 min). Score against the rubric below, then do one rewrite of your weakest line.
Subject: onboarding load + your backend hiring Hi Priya, Saw Acme's hiring six backend engineers this quarter - usually that means onboarding and PR review time both spike right when the team can least afford it. A few teams your size use Cursor to cut ramp time on big TypeScript codebases - new hires get productive faster because the editor understands the repo, not just the file. Worth 20 min Thursday to see if it'd actually help your team or is that not a priority right now? - Alex
Grade your draft on the SDR-writing rubric
- Dimension
- Personalization
- A 5 looks like
- A specific, real trigger only this account has
- A 2 looks like
- "I came across your company" - works for anyone
- Dimension
- Clarity
- A 5 looks like
- Under 90 words, one idea, scannable on a phone
- A 2 looks like
- Three paragraphs, two asks, buried point
- Dimension
- No-hype framing
- A 5 looks like
- Plain claim an engineer would believe; outcome, not adjectives
- A 2 looks like
- "Revolutionary AI-powered 10x productivity"
- Dimension
- CTA
- A 5 looks like
- One specific, low-friction ask with an easy out
- A 2 looks like
- "Let me know if you'd ever want to connect sometime"
| Dimension | A 5 looks like | A 2 looks like |
|---|---|---|
| Personalization | A specific, real trigger only this account has | "I came across your company" - works for anyone |
| Clarity | Under 90 words, one idea, scannable on a phone | Three paragraphs, two asks, buried point |
| No-hype framing | Plain claim an engineer would believe; outcome, not adjectives | "Revolutionary AI-powered 10x productivity" |
| CTA | One specific, low-friction ask with an easy out | "Let me know if you'd ever want to connect sometime" |
Anything under 4 on a row = rewrite that line before you stop.
Hype is poison with engineers. "Revolutionary," "transformative," and "10x" all read as someone who has never shipped code. Cursor sells to skeptics, so a plain outcome they can picture - "new hires get productive faster because the editor understands the repo" - beats every superlative you could reach for.
Give the buyer a graceful out in your CTA: "...or is that not a priority right now?" Naming the no makes the yes easier and signals you respect their time. It reads as confident, not desperate, which is exactly the tone a technical buyer responds to.
Takeaway. A strong take-home is short, trigger-led, hype-free and ends in one specific CTA - proofread to zero typos because it is Cursor's first impression of you.
Self-check
QWhich cold-email opener is strongest for an engineering persona at a target account?
The scorecard and gap plan
After this you can After this you can score yourself to the real rubric and build a focused fix list.
Reps without a scorecard are just vibes. Grade each stage on its own rubric, find your two weakest categories and drill only those - short daily reps beat a single cramming marathon.
Score every stage 1–5 on the dimensions that stage tests. The number itself matters less than catching the same weakness twice across reps. If your close keeps scoring a 2, that is your work, not your nerves.
Interactive diagram. Tab through its regions; each focused region shows its detail in the panel below.
Weight your scorecard toward the signals the panel actually weighs most.
The full-loop scorecardScore each row 1–5 after every rep
- Stage
- Recruiter / HM screen
- What you're scoring
- Why-Cursor specificity, pipeline math, AE-path conviction, conciseness
- Pass bar (4+)
- Crisp, product-rooted, numbers with ratios, no filler
- Stage
- Cold call opener
- What you're scoring
- Permission-based, honest, earns the next 30 seconds
- Pass bar (4+)
- Buyer grants time instead of hanging up
- Stage
- Discovery
- What you're scoring
- Talk-time ~80% them, real questions, quantified pain
- Pass bar (4+)
- You learned their workflow; you didn't pitch
- Stage
- Objection handling
- What you're scoring
- Acknowledge–Ask–Advance, calm, never argues
- Pass bar (4+)
- Each objection reopened the conversation
- Stage
- Close
- What you're scoring
- Specific meeting ask with name + slot + an out
- Pass bar (4+)
- A concrete next step, not vague time
- Stage
- Written take-home
- What you're scoring
- Personalization, clarity, no hype, one CTA, zero typos
- Pass bar (4+)
- Could only have been written for that account
- Stage
- Values / why-Cursor
- What you're scoring
- Truth-seeking, ownership, real product passion
- Pass bar (4+)
- Conviction holds up under follow-up probing
| Stage | What you're scoring | Pass bar (4+) |
|---|---|---|
| Recruiter / HM screen | Why-Cursor specificity, pipeline math, AE-path conviction, conciseness | Crisp, product-rooted, numbers with ratios, no filler |
| Cold call opener | Permission-based, honest, earns the next 30 seconds | Buyer grants time instead of hanging up |
| Discovery | Talk-time ~80% them, real questions, quantified pain | You learned their workflow; you didn't pitch |
| Objection handling | Acknowledge–Ask–Advance, calm, never argues | Each objection reopened the conversation |
| Close | Specific meeting ask with name + slot + an out | A concrete next step, not vague time |
| Written take-home | Personalization, clarity, no hype, one CTA, zero typos | Could only have been written for that account |
| Values / why-Cursor | Truth-seeking, ownership, real product passion | Conviction holds up under follow-up probing |
Average is a vanity metric; your two lowest rows are the plan.
Turn the score into a drill plan
- Pick
- Your two lowest-scoring categories only
- Drill
- 10–15 min/day on those, not the whole loop
- Rerun
- Only the failed stage, against a fresh persona
- Track
- Same scorecard each rep; watch the number move
- Stop
- When the weak rows hold at 4+ across two reps
You are ready when you would hire yourself. Run the full loop, score it honestly and ask: if I were the hiring manager and this candidate showed up, would I move them forward without hesitation? If the answer is yes for every stage, you are ready. If you flinch on one row, that row is your last week of work.
Do not re-run the whole loop every time - it feels productive and it isn't. If your call scored a 5 and your writing scored a 2, running the call again just rehearses a strength while your real gap stays a gap. Re-run the failed stage only and spend the saved time on the two rows that are actually holding you back.
Takeaway. Score each stage on its own rubric, drill only your two weakest rows with short daily reps and call yourself ready only when you'd hire yourself.
Self-check
QYour cold call scores a 5 but your written take-home scores a 2. What's the most effective next move?
Final-week game plan and day-of execution
After this you can After this you can walk in prepared, calm and unmistakably a hire.
The last week is not for cramming new material. It is for sharpening what you have, locking your story bank and using Cursor every single day so your product passion is current, not remembered.
By now your gaps are known. The final week converts a prepared candidate into a calm one. Light, frequent reps keep you warm without burning you out and daily product use keeps your "why Cursor" honest.
The final-week planTaper, don't cram
- 1Refine the story bank. Tighten your self-sourcing story, your resilience story and your honest product-usage story so each runs clean in under 90 seconds.
- 2Do 3–5 short call reps. Five focused minutes a day keeps your opener and objection handling warm. You are maintaining, not rebuilding.
- 3Lock your "why Cursor." One version, memorized to the bone, that could only be said about Cursor. This is the trait they screen hardest for.
- 4Use the product daily. Ship something small in Cursor each day so you can speak from this week, not last month - "I used the agent yesterday to..." lands far harder than a feature list.
Logistics to lock before the day
- Confirm SF / onsite details: address, who you're meeting, arrival time and buffer for traffic.
- Tech-check any video stage: camera, mic, lighting and a backup connection.
- Prepare a short, real question list per round - sharper questions for the AE and founder than for the recruiter.
- Have your researched take-home account and notes ready, in case the AE round reviews it live.
Day-of framing to yourself before each round: "This is a conversation between two people figuring out if this is a fit - not an interrogation I have to survive." Warm your voice up out loud beforehand, take one slow breath before you start and let the prep carry you.
Treat your thank-you note as another writing sample, because it is. Keep it short, reference one specific thing from the conversation and proofread it as hard as the take-home. A concise, well-written follow-up reinforces the exact "clear, concise writing" the role demands; a sloppy one quietly undoes a good loop.
Above every framework, Cursor wants conviction about automating coding. The candidate who genuinely loves the product and the problem beats the smoother candidate who treats it as just a job - its CEO has said as much. Bring that passion truthfully and let it show; it is the single trait they most want to see and it cannot be faked under follow-up.
Takeaway. Final week is taper, not cram: lock your "why Cursor," keep reps short, use the product daily and walk in treating each round as a conversation - with conviction as your closing trait.