No card - 3 minutes

Take a door.
Right now.

Three minutes against an AI homeowner who argues, stalls, and shuts the door. Your first name, your mic, and you’re on the porch. The score lands on screen the second you hang up.

Hear him first
Hostile cold knock · sample
0:20
Step 1 of 3 · name → mic → knock

The homeowner will call you by it.

Mic required · 3-minute cap · no card, no account

01The loop

Knock. Score. Again.

Drill length depends on the door: a cold knock or warm lead caps at 15 minutes and usually runs about 5, a booked sit-down caps at 40. The loop is the product: reps repeat the exact door they just lost until they stop losing it.

01

Pick the door

Cold knock, warm lead, or booked sit. Dial difficulty from easy to hard and temperament from calm to hostile: nine versions of every door. Each type runs to its own cap: 15 minutes on a cold knock or warm lead, 40 on a booked sit-down.

02

Knock and talk, live

A real spoken conversation over your mic, in the browser. The homeowner interrupts, deflects, and bites back the way Saturday-morning doors actually do. Most door drills are over in about 5 minutes. The cap is a ceiling, not a target.

03

Get scored. Run it again.

End the call and the debrief lands, opening with the one thing that cost you the door, then a band across seven dimensions and the top 3 corrections with the line to say instead.

02The debrief

Every call ends in ink.

End the call and an LLM reads the whole transcript against a fixed 7-dimension rubric. Discovery weighs heaviest: the sale gets made in the questions, not the presentation.

No participation ribbon. It opens on the one thing that cost the rep the door, then the band, then three corrections quoted at the line level. Run the same door tomorrow and beat it.

This is the real component with a sample drill in it, not a mockup. The dimension labels and the band come from the same rubric the app renders from.

What cost you this door
You answered the price question before you knew what his bill was. From there he was shopping you, not listening.
Solid

Competent. Worked the door, left something on the table.

DiscoveryWeighted
3/5
Objection handling
4/5
Acknowledged the no
3/5
Opener / pattern break
4/5
Pressure control
3/5
Outcome & disengage
4/5
Word choice & phrasing
4/5
What worked

You took the first no without flinching and kept it neighbor-level. The opener earned a real question inside 20 seconds.

Top 3 corrections
  1. 01You pitched panels before asking a single question. Open the bill first: “Out of curiosity, what’s electric running you in summer?”
  2. 02When he said “we already got quotes,” you argued price. Anchor the comparison instead: “Same equipment and same warranty, or just a lower number?”
  3. 03You held the door past the second no. Take it clean and set the exit: “Fair enough. Last thing and I’m gone.”
How to read this

The grader reads the transcript. It sees your words and your order of operations, not your pace or your warmth. The band, not a mark out of 100, is the resolution it actually has.

Fig. 02. The debrief, as the app renders it. Sample drill: cold knock, hard, hostile.
03The rubric

Seven ways a call gets graded.

Every debrief scores these seven, weighted the way the door rewards them: discovery over everything, phrasing last. The weights are a judgement call, published so you can argue with them: they have not been validated against closed deals, and we won't say they have.

30
weight
DiscoveryWeighted heaviest01

Did you dig into their bill, roof, and real motivation before you pitched a thing? The heaviest line on the card: the sale gets made in the questions, not the presentation. Question COUNT is not the goal: at a real door the fourth question from a stranger is when the storm door closes, and the grader is told to score whether the questions were earned and landed, not how many you fired.

20
weight
Objection handling02

Did you answer the actual objection with specifics, or wave it off and talk over it?

15
weight
Acknowledged the no03

Did you respect the refusal out loud before pushing past it? Skip this and the door shuts on you. A text grader can be gamed here ("I hear you, and I respect that" is a phrase, not a behaviour), so the grader is told to credit this only when the rep's NEXT move actually respects the no.

10
weight
Opener / pattern break04

Did you survive the reflex “not interested” without ignoring it or bulldozing?

10
weight
Pressure control05

Did you avoid fake urgency and the premature same-day close?

10
weight
Outcome & disengage06

Did you earn a soft next step, or exit clean without burning the door?

5
weight
Word choice & phrasing07

Did you talk like a person: plain words, short turns, no script on legs? Scored from the transcript only. Pace, pauses and warmth aren't in a transcript, so we don't pretend to grade them.

What the grader sees

An LLM reads the full transcript of the drill against these seven, in about 30 seconds. It reads words, so it grades what the rep said and in what order. It does not hear the call, which is why the seventh line is phrasing rather than tone: pace, pauses, volume and interruption timing are not in a transcript and nothing here claims to score them.

The weighted total is kept as a number so a rep’s curve can be tracked over weeks, but the debrief shows a band, not a point score. A model asked for a mark out of 100 will give you one; it does not actually separate a 73 from a 76.

04The math

Run it against your own numbers.

Your rep's alternative is knocking one more street, and it's free. So do the arithmetic with your gross, your close rate, your headcount. Everything below is yours to change, and none of it is a claim we're making about what this tool does.

Your numbers
Plan you’d be on
Where it washes
0.04

extra sits per rep per month is what Team at $349 costs you, at your gross and your close rate. Below that it loses money. Above it, it doesn’t. Whether the drills move that number is the thing you should test, not take our word for.

6.0
Extra installs / mo
at 2 extra sits per rep
$18,000
Extra gross / mo
vs $349 for the plan
One extra deal per rep, per quarter

If every rep closed one more job a quarter and nothing else changed, that’s $10,000 of gross a month across 10 reps, against $349 for Team. That comparison is the one worth arguing about; it’s also the one we can’t make for you.

Every figure here is computed from the fields on the left. CloseDrill has no published outcome data and this calculator does not stand in for any. It’s your arithmetic, run faster.

05Pricing

What it costs.

7-day free trial on Solo and Team. Annual pays for 10 months and gets 12.

Solo1 seat
$79/mo

One closer sharpening before the field.

  • +All 3 door scenarios
  • +Full difficulty × temperament matrix
  • +Scored debrief on every call
  • +Personal call history
Start on Solo
Most popular
Teamup to 10 seats
$349/mo

The core plan: a solar sales team.

  • +Everything in Solo
  • +Seats for up to 10 reps
  • +Rep invites by company code
  • +Your company name in-app
Start on Team
Companyup to 30 seats
$899/mo

Larger installers running a bench.

  • +Everything in Team
  • +Seats for up to 30 reps
  • +Manager dashboard
  • +Team leaderboard
Start on Company
Annual: Solo $790 · Team $3,490 · Company $8,990Access code? It applies at checkout, and can cover the plan in full.
06Questions

Straight answers.

Does it use my real phone?

No. Drills run in the browser over your mic: laptop or phone, truck or office. Nothing dials out, and no homeowner ever hears a practice rep.

What do the calls cost me?

Not unlimited. Anyone telling you voice minutes are free is paying for them somewhere, and we'd rather say so than surprise you. Every drill is hard-capped by type: 15 minutes on the door scenarios (most run about 5), up to 40 on a full booked-appointment sim. Beyond those caps, minutes are metered against your plan rather than absorbed. If you're putting a large team on this and expect heavy daily use, ask us where your allowance sits before you sign. That's a real conversation, not a gotcha.

How long does a drill take?

It depends on the door. A cold knock or warm lead caps at 15 minutes and usually runs about 5. You hang up when you've got the sit or lost it. A booked-appointment sim is a full sit-down and caps at 40. The free demo on the homepage is a cold knock capped at 3 minutes.

How is the scoring done?

An LLM reads the full transcript against a fixed 7-dimension rubric: discovery (weighted heaviest), objection handling, acknowledging the no, the opener, pressure control, outcome and disengage, and word choice. It reads the transcript, not the audio, so it grades what the rep said, not how they sounded; pace and tone are not scored, because they aren't in a transcript. The rep sees a band rather than a mark out of 100, since a model asked for a number will give you one but doesn't really separate a 73 from a 76. Same rubric every call, so bands compare across reps and across weeks. The weights are our judgement call, published in full, and not yet validated against closed deals.

Do my reps need accounts?

Yes. On Team and Company, a manager generates an invite code and reps join under the company. About a minute per rep.

Can I white-label it for my company?

Your company name shows in-app for your reps. Full white-label isn’t offered.

I was handed an access code. What does it do?

Some partner companies get CloseDrill included with other services. The code applies at checkout like a coupon and can zero out the plan. Redeem it at sign-in.

The next knock is free.

Run a cold door against a hostile homeowner in the next ten minutes. Seven days free on Solo and Team. No leads burned while your reps learn.

Take a door - 3 min