Benchmark rows

# Not live yet (gaps #2)
curl -X POST "https://api.sb/leaseline.app/campaigns?mode=try" \
  -H "Authorization: Bearer $STARTUP_BUILDER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"play":"benchmark-rows"}'

 1import { $ } from 'startup-builder'
 2
 3return await $.Campaign.create({ play: 'benchmark-rows' }, { mode: 'try' })   // dry run on the day-0 Startup, 2026-08-24 · Not live yet, gaps #2

npx startup-builder do Campaign.create --json '{"play":"benchmark-rows"}' --try   # Not live yet (gaps #2)

{
  "name": "do",
  "arguments": {
    "code": "return await $.Campaign.create({ play: 'benchmark-rows' }, { mode: 'try' })   // dry run on the day-0 Startup, 2026-08-24 · Not live yet, gaps #2",
    "mode": "try"
  }
}

 1{
 2  "type": "https://startupbuilder.dev/errors/refused",
 3  "title": "Refused",
 4  "status": 422,
 5  "detail": "benchmark-rows can’t start yet: 2 checks fail and 1 warns.",
 6  "checks": [
 7    {
 8      "check": "owned-by-a-person",
 9      "status": "pass",
10      "observed": "0 of 640 companies have a Lead or open Deal a person owns",
11      "doc": "/plays/benchmark-rows#check-owned-by-a-person",
12      "then": "Task.assign"
13    },
14    {
15  …

curl "https://api.sb/leaseline.app/events?where=name%20in%20(%22Monitor.checked%22%2C%20%22Form.submitted%22)&limit=1" \
  -H "Authorization: Bearer $STARTUP_BUILDER_API_KEY"

# Not live yet (gaps #7)
curl -X POST "https://api.sb/leaseline.app/triggers/fire?mode=try" \
  -H "Authorization: Bearer $STARTUP_BUILDER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"like":"event_1423"}'

# Not live yet (gaps #7)
curl "https://api.sb/leaseline.app/decisions?where=on%20%3D%20%22organization_144%22" \
  -H "Authorization: Bearer $STARTUP_BUILDER_API_KEY"

# Not live yet (gaps #7)
curl "https://api.sb/leaseline.app/actions?where=on%20%3D%20%22organization_144%22" \
  -H "Authorization: Bearer $STARTUP_BUILDER_API_KEY"

 1import { $ } from 'startup-builder'
 2
 3const { events: [last] } = await $.Event.list({ where: 'name in ("Monitor.checked", "Form.submitted")', limit: 1 })   // { play } compiled to this Play’s Trigger Events
 4const { subject } = await $.Trigger.fire({ like: last.$id }, { mode: 'try' })   // the same Event again, in try mode · Not live yet, gaps #7
 5const { decisions: [decision] } = await $.Decision.list({ where: `on = "${subject}"` })   // what it chose, and why · Not live yet, gaps #7
 6const { actions } = await $.Action.list({ where: `on = "${subject}"` })   // what it did, with its checks · Not live yet, gaps #7
 7return { decision, actions }

npx startup-builder do Event.list --json '{"where":"name in (\"Monitor.checked\", \"Form.submitted\")","limit":1}'
npx startup-builder do Trigger.fire --json '{"like":"event_1423"}' --try   # Not live yet (gaps #7)
npx startup-builder do Decision.list --json '{"where":"on = \"organization_144\""}'   # Not live yet (gaps #7)
npx startup-builder do Action.list --json '{"where":"on = \"organization_144\""}'   # Not live yet (gaps #7)

{
  "name": "do",
  "arguments": {
    "code": "const { events: [last] } = await $.Event.list({ where: 'name in (\"Monitor.checked\", \"Form.submitted\")', limit: 1 })   // { play } compiled to this Play’s Trigger Events\nconst { subject } = await $.Trigger.fire({ like: last.$id }, { mode: 'try' })   // the same Event again, in try mode · Not live yet, gaps #7\nconst { decisions: [decision] } = await $.Decision.list({ where: `on = \"${subject}\"` })   // what it chose, and why · Not live yet, gaps #7\nconst { actions } = await $.Action.list({ where: `on = \"${subject}\"` })   // what it did, with its checks · Not live yet, gaps #7\nreturn { decision, actions }",
    "mode": "try"
  }
}

 1{
 2  "decision": {
 3    "$id": "https://api.sb/leaseline.app/decisions/decision_495",
 4    "$type": "Decision",
 5    "choice": "send-the-row",
 6    "reason": "Brookstone Rentals scores 41 against the median 58; its renewals page names no notice date, a fix it can copy from the method page, and its operations lead has a verified address",
 7    "stakes": "low",
 8    "options": [
 9      "send-the-row",
10      "send-as-you",
11      "ask-to-be-named",
12      "send-the-movement",
13      "re-read-it",
14      "not-now"
15  …

action_3 is the row note to Brookstone’s operations lead, held 2 minutes, with a link to its private row page; the row page is noindex, and only Brookstone sees its own score.

≈ +$250 ARR a month at 95 rows

Worth

1 h week 1, then 20 min a week

Your time

$0.06 per row sent, ~$6 a month

Spend

Day 5: first company ships its fix

First signal

300+ companies with public pages, a free fix to show

Use it when

Under 300: proof-of-work-outreach · no free fix: cold-email

Not yet when

Worth’s formula: about +7.9 acting companies per 100 rows × 95 rows a month × 0.43 actions to Leads × 0.12 Leads to paid × $650 first order ARR ≈ +$250. Under First signal: About 4 months to tell a +8 lift at 95 rows a month with 10 % held back; at 300 companies, the 30-day read pools with Other Startups’ results.

Runs inside: Proof-of-work outreach, as one of its options; this page keeps its own record and URL.

Start it this weekCopy Anchor Link

npx startup-builder add play benchmark-rows   # 1 · writes plays/benchmark-rows.mdx (Not live yet, gaps #5)

# 2 · dry run; below, its 422 on the day-0 Startup (clock at 2026-08-24)
curl -X POST "https://api.sb/leaseline.app/campaigns?mode=try" \
  -H "Authorization: Bearer $STARTUP_BUILDER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"play":"benchmark-rows"}'

# 3 · keep it, at You approve each one for the first 25 rows
curl -X POST https://api.sb/leaseline.app/campaigns \
  -H "Authorization: Bearer $STARTUP_BUILDER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"play":"benchmark-rows"}'

# on
curl -X POST https://api.sb/leaseline.app/processes/outbound/enable \
  -H "Authorization: Bearer $STARTUP_BUILDER_API_KEY"

# off: no row note sends; the index stays published
curl -X POST https://api.sb/leaseline.app/processes/outbound/pause \
  -H "Authorization: Bearer $STARTUP_BUILDER_API_KEY"

{
  "type": "https://startupbuilder.dev/errors/refused",
  "title": "Refused",
  "status": 422,
  "detail": "benchmark-rows can’t start yet: 2 checks fail and 1 warns.",
  "checks": [
    {
      "check": "owned-by-a-person",
      "status": "pass",
      "observed": "0 of 640 companies have a Lead or open Deal a person owns",
      "doc": "/plays/benchmark-rows#check-owned-by-a-person",
      "then": "Task.assign"
    },
    {
      "check": "use-when",
      "status": "pass",
      "observed": "640 property managers with public renewal pages",
      "doc": "/plays/benchmark-rows#check-use-when",
      "then": "Metric.query"
    },
    {
      "check": "baseline",
      "status": "warn",
      "observed": "0.8 hand-raises per 100 of these companies in 90 days, n 640",
      "doc": "/plays/benchmark-rows#check-baseline",
      "fix": "n 640: the held-back 10 % is your comparison from day 1.",
      "then": "Metric.query"
    },
    {
      "check": "approved-pages",
      "status": "fail",
      "doc": "/plays/benchmark-rows#check-approved-pages",
      "fix": "Publish the method page; every row links to it.",
      "then": "Site.publish"
    },
    {
      "check": "domain-authenticated",
      "status": "pass",
      "observed": "leaseline.app: 3 of 3 records resolve",
      "doc": "/plays/benchmark-rows#check-domain-authenticated",
      "then": "Domain.authenticate"
    },
    {
      "check": "mailbox-warm",
      "status": "fail",
      "doc": "/plays/benchmark-rows#check-mailbox-warm",
      "fix": "Wait for the warm-up to finish, 14 days from day 0; the index publishes meanwhile.",
      "then": "Mailbox.warm"
    },
    {
      "check": "role-unstaffed",
      "status": "warn",
      "roles": [
        "sales-lead"
      ],
      "owned": 0,
      "doc": "/plays/benchmark-rows#check-role-unstaffed",
      "fix": "Staff the AE, or leave it: replies come to you.",
      "then": "Role.staff"
    }
  ]
}

Each reading costs $0.01 a page, at most 10 pages a company; each row note costs $0.04 from your Domain (Worker cost), and contact data for the role that owns the fix costs $0.02 (POST /{namespace}/contacts/{contact}/enrich). That comes to about $6 a month at 95 rows and a weekly re-read.

When it startsCopy Anchor Link

{
  "$id": "https://api.sb/leaseline.app/events/event_1423",
  "$type": "Event",
  "name": "Monitor.checked",
  "at": "2026-10-19T06:30:19Z",
  "seq": 1423,
  "data": {
    "monitor": "renewal-readiness",
    "organization": {
      "$id": "https://api.sb/leaseline.app/organizations/organization_144",
      "name": "Brookstone Rentals"
    },
    "score": 41,
    "median": 58,
    "gap": 17,
    "fix": "state the notice date on the renewals page"
  }
}

# each delivery is signed with the Watch’s secret
curl -X POST https://api.sb/leaseline.app/watches \
  -H "Authorization: Bearer $STARTUP_BUILDER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"selector":"Monitor.checked","url":"https://leaseline.app/hooks/sb"}'
npx startup-builder trigger Monitor.checked --like event_1423   # replays the ready row
npx startup-builder clock advance --by 7d   # runs the weekly re-read, which finds a fix if it shipped

Who runs itCopy Anchor Link

SDR · your agent (role:prospector, worker_2) holds the Key Result. It hands replies to AE · Priya Shah (person) (role:sales-lead, worker_3, staffed on day 9), and top-tier rows, re-reads that change a score and the top 10 to you (Dana Okafor, person, worker_1). Your number links to Did it work.

What it reads · What it may call. It reads public pages only, at a rate Limit, and stores each reading with its page and time. A company’s own row goes only to that company; public text names only a top-10 company that said yes. Its tools are POST /{namespace}/messages/{message}/schedule, POST /{namespace}/monitors/{monitor}/check, and any other call becomes an Approval under the Role’s Mandate.

You hold the SDR Key Result: 15 acting companies per 100 rows by 2026-12-31.     ← worker, metric
On each ready row, re-read or request, choose one option and say why.            ← triggers, chooses
Every number in a note matches a stored reading under 7 days old.                ← number-match, fresh-reading
Every row note gives one fix the company can make for free.                      ← your file, line 4; fix-gate
Name nobody outside the top 10, and them only with their yes.                    ← your file, line 5; named-score

The slow loop. The weekly Decisions kept the 8-point gap floor and the one-fix rule, and no Change was merged in the window, so there is no replay or read-back to print.

Bring your own agent (folded) starts at You approve each one when it trails the seated agent on unmatched numbers, and at It does it, you can stop it when it matches.

npx startup-builder eval --scenario benchmark-rows-191 --agent "claude -p" --seat prospector   # seated: 13.8 acting per 100, 0 unmatched numbers, $0.06 a row · yours: 12.9, 2 unmatched → You approve each one

What it can do nextCopy Anchor Link

When it hands to youCopy Anchor Link

Replies go to AE · Priya Shah (person); top-tier rows, the top 10 and any re-read that changes a score come to you. Before day 9 every reply came to you. What lands in each queue, as captured: one Brief per Role, then the Approval (brief, draft, route and checks are Not live yet, gaps #7).

[
  {
    "$id": "https://api.sb/leaseline.app/tasks/task_297",
    "$type": "Task",
    "assignee": "worker_3",
    "due": "2026-09-25T15:04:00Z",
    "on": "lead_72",
    "brief": {
      "who": "Mei Tanaka, Northgate Lettings",
      "said": "Fixed the notice date. What else would move us?",
      "chose": "send-the-row",
      "why": "scored 49 against 58; shipped the fix in 6 days",
      "row": "61 after the re-read, up 21 percentiles",
      "never": [
        "another company’s score",
        "a price beyond the list"
      ]
    }
  },
  {
    "$id": "https://api.sb/leaseline.app/tasks/task_301",
    "$type": "Task",
    "assignee": "worker_1",
    "due": "2026-10-02T17:00:00Z",
    "on": "organization_151",
    "brief": {
      "who": "Harlow Lettings",
      "said": "Your reading of our fees page is wrong.",
      "chose": "re-read-it",
      "why": "a disputed reading",
      "reread": "2 reads 24 h apart: 44 and 52; their fees page changed on Sep 29"
    }
  },
  {
    "$id": "https://api.sb/leaseline.app/approvals/approval_86",
    "$type": "Approval",
    "verdict": null,
    "deadline": "2026-10-02T17:00:00Z",
    "draft": "Harlow’s row moves from 44 to 52 after the Sep 29 change to its fees page.",
    "route": "re-read-it",
    "checks": [
      {
        "check": "number-match",
        "status": "pass"
      },
      {
        "check": "fresh-reading",
        "status": "pass"
      }
    ]
  }
]

Each receiver answers with POST /{namespace}/tasks/{task}/claim (Take this one), POST /{namespace}/tasks/{task}/return with a reason, POST /{namespace}/actions/{action}/redirect, POST /{namespace}/approvals/{approval}/approve or POST /{namespace}/approvals/{approval}/refuse, each a one-call window, and an Approval unanswered by its deadline expires and the score stays as read. Once a person claims, the row notes to that company stop, and every reply goes to the claimer’s queue. A Task nobody claims by its due time goes up a Role through POST /{namespace}/tasks/{task}/escalate (AE to you), and person-promise fires.

On time: AE · Priya Shah, 8 of 8 replies claimed within 60 min, median 26 min, ■■□□ Early sign, n 8; you, 17 of 17 approvals by their deadline, n 17. Your queue: counted by kind: in weeks 1 to 3, 25 row approvals, the method page and 10 top-10 asks; since, about 1 hand-off and 1 re-read a week.

The fixed line: POST /{namespace}/deals/{deal}/assign sets owner to the receiving person, the agent is never a Deal’s owner, and Meetings land on the owner’s calendar under the owner’s name.

When it won’t sendCopy Anchor Link

The captured refusal, then the Stop that fired as a Campaign.paused reason (Not live yet, gaps #2); the note that sent instead is the second thread in Example.

{
  "type": "https://startupbuilder.dev/errors/refused",
  "title": "Refused",
  "status": 422,
  "detail": "The row note to Oak Hill Management did not send: it states a median the readings do not hold.",
  "checks": [
    {
      "check": "number-match",
      "status": "fail",
      "observed": "the note said 60; the stored median is 58",
      "doc": "/plays/benchmark-rows#check-number-match",
      "fix": "Use the stored reading, or re-read the page first.",
      "then": "Monitor.check"
    },
    {
      "check": "fresh-reading",
      "status": "pass"
    }
  ]
}

[
  {
    "reason": {
      "limit": "sending-stop-loss",
      "observed": "2.3 % bounces on rows@leaseline.app",
      "bound": "2 %",
      "window": "1 day"
    },
    "doc": "/plays/benchmark-rows#stop-sending-stop-loss",
    "mailbox": "rows@leaseline.app"
  }
]

Did it workCopy Anchor Link

Both halves come from the same random draw: 191 companies sent their row and 21 held back (rules m–o); a company counts once 30 days have passed since its row. Each figure’s call is POST /{namespace}/metrics/query by node and by arm; every node above the leaf is Not live yet (gaps #4).

{
  "$id": "decisions/decision_495",
  "$type": "Decision",
  "choice": "send-the-row",
  "reason": "Brookstone Rentals scores 41 against the median 58; its renewals page names no notice date, a fix it can copy from the method page, and its operations lead has a verified address",
  "stakes": "low",
  "options": [
    "send-the-row",
    "send-as-you",
    "ask-to-be-named",
    "send-the-movement",
    "re-read-it",
    "not-now"
  ],
  "by": "worker_2",
  "on": "organization_144",
  "expected": "Brookstone ships the fix, replies or claims its row within 30d",
  "reopensWhen": "no row action at organization_144 within 30d"
}

The leaf counts companies that, within 30 days of their row, ship a change that moves their score, reply with anything but stop, or claim their row, each once, per 100 rows delivered; a held-back company counts if it writes in first.

How it moves it: each company got its own score, its gap to the median of 58 and one fix it could copy from the method page, and 14 of the first 152 shipped that fix within 30 days.

Guardrails: treated against held back: positive-reply-rate on the row Mailboxes, 6.1 %, and removal asks, 2 of 188 delivered, under the 3 % Stop.

{
  "$id": "https://api.sb/leaseline.app/campaigns/benchmark-rows",
  "$type": "Campaign",
  "play": "play:benchmark-rows@1",
  "path": [
    "row-action-rate",
    "new-leads",
    "new-arr",
    "net-new-arr",
    "arr"
  ],
  "experiment": {
    "metric": "row-action-rate",
    "unit": "per-100",
    "heldBack": 21,
    "decision": null
  },
  "readout": {
    "value": 7.9,
    "n": 169,
    "maturing": 43,
    "interval": [
      -4.6,
      20.4
    ],
    "readability": "inadmissible",
    "basis": "held-back",
    "pooled": false
  },
  "ledger": {
    "spend": 9,
    "workerCost": 8,
    "hours": 3.1,
    "wonArr": 945,
    "projectedArr": 250,
    "paybackDays": 7
  }
}

Outside benchmark: self-ranked “best of” lists leave out their own publisher in 1 of 4 cases, and readers trust those that do (a 2026 search-quality study).

Before it · After it. Before it: programmatic-seo and aeo make the index page findable. After it: proof-of-work-outreach, which it runs inside, writes the next note once a company engages, and instant-response answers anyone who writes first.

ExampleCopy Anchor Link

Leaseline ran it from 2026-08-24 to 2026-10-19 (56 days). 640 property managers were read; 26 could not be (4.1 %, under the hold), leaving 614 scored, median 58. After the top 10, 31 Customers, no fix, no address and holds, 212 were eligible: 191 got their row and 21 were held back at random. Read every captured thread at ?threads=all; three follow the tables.

Leaseline → Mei Tanaka, Northgate Lettings, Sep 12, 08:40 · Subject: Northgate’s renewal readiness: 49 · We read 614 property managers’ public renewal pages; the median is 58.¹ Northgate’s renewals page names no notice date, which is 9 of the 51 points between you and the leader. Adding one line, “we send renewal notices 60 days out”, closes it.² Your private row: leaseline.app/rows/northgate. We run this index and sell renewal software. *Signed with the Role’s disclosure line.*³

Sep 18 · the re-read found the line live: 61, up 21 percentiles; the movement note went the next morning. Sep 25 · “Fixed the notice date. What else would move us?” The Brief went to Priya, and Northgate took Team on Oct 9 (+$945 ARR).

- The median property manager scores 60; Oak Hill scores 47.
+ The median property manager scores 58; Oak Hill scores 47.

Harlow Lettings, 2026-10-01 11:05 · “Your reading of our fees page is wrong.” · Leaseline, 11:06 · *Thanks. Dana re-reads every disputed row and replies by tomorrow, 17:00.*⁴ Two reads 24 h apart scored 52, up from 44 after their Sep 29 change; Dana approved the new score at 15:30 (approval_86).

¹ Every number matches a stored reading under 7 days old. ² One fix the company can make for free. ³ The fixed signature and steward lines, compiled from the Role’s Settings. ⁴ The routed line, compiled from the Role’s Settings.

Your settingsCopy Anchor Link

---
holdout: 0.1      # 1 in 10 eligible companies gets no row note, chosen at random
---
Score renewal readiness, 0 to 100, from public pages only; one free fix per row.
Name only the top 10, and only with their yes.
Over 500 units, send the row as Dana, after her OK.

$ npx startup-builder show plays/benchmark-rows
holdout 0.1 · the index and fix rule · the top-10 rule · the 500-unit line    your file  plays/benchmark-rows.mdx:2,4,5,6 (Ask 2026-08-24)
family demand · process outbound · runs inside proof-of-work-outreach · worker role:prospector, owns never · chooses 6 (4 rule, 2 read) · sequences row-first-touch, row-first-touch-founder, row-movement, row-watch, top-ten-ask · approval every-send · checks owned-by-a-person, number-match, fresh-reading, fix-gate, named-score, cell-floor, self-exclusion, steward-line, method, movement-limit, late-note · stops 7    catalog  play:benchmark-rows@1 (Not live yet, gaps #5)

Every key from the catalogCopy Anchor Link

Compiled, not written: edit the Play record benchmark-rows and its Leaseline run, then regenerate. Every step runs from a sentence or a line of code. In your words: the CRM words you know, and where each lives here.