Run log summary for the simulated night shift, 23:41 to 06:00: the shift armed at 23:41. Lights out at 00:00; five modules online at 00:30. Revenue reconciliation (REC-01) tied the ledger at 01:12 — delta 0.4%, the pending T+2 batch explained, and the morning's missing $340 found in the Shopee batch settled at 00:47. Ad anomaly detection (SIG-02) caught a +212% CPM spike at 03:14, scoring 4.9 times the median absolute deviation, and paused the ad set under the operator's threshold rule. Inventory velocity (INV-04) flagged SKU-104 at 04:02 — 11 days of cover against a 14-day lead. Review sentiment (REV-05) flagged the phrase theme "sizing runs small", 31 occurrences over 7 days, at 04:47. The Owner Brief was written and sent at 05:30. Shift complete at 06:00: 4 signals resolved, 1 brief written, 0 humans woken. All figures simulated.

SYS-01 // ARMED 23:41

LIGHTS OUT

Lights out. So you can sleep.

Five automations work the night. One brief by 06:00.

ONE SHIFT · 23:41 → 06:00 · SCROLL RUNS THE CLOCK · FOR DTC BRANDS $30K–$500K/MO

ARCHIVE REPLAY — TODAY

THE GLARE

  1. 07:00You open Meta, eyeball yesterday's spend, wince.
  2. 07:15You pull the Shopify CSV. Shopee reports in a different currency.
  3. 07:30The numbers don't tie. $340 is missing. You have a 9am, so you move on.You move on.
  4. 14:00A bundle SKU stocks out. Supplier lead is 14 days. Your best-margin SKU, dark for 2 weeks.
  5. 23:00You're still in the spreadsheet, working out which channel paid for itself.
FIG. A — THE MANUAL PASS

WRONG BELIEF

"You need better ads to scale."

The brands that scale aren't out-creating their competitors. They're out-operating them.

The business runs on you being awake. That is the bug.

EVERY MORNING
45 MIN
EVERY WEEK
6 DAYS
ACCEPTED ERROR
2–5%

That gap is where every budget decision gets made wrong for the rest of the day.

00:00

SYS-01 · LIGHTS OUT · STATUS to RUNNING

// 00:30

THE FLOOR

  • LEDGER — ONE SOURCE
  • REC-01 — RECONCILIATION
  • SIG-02 — ANOMALY DETECTION
  • BRF-03 — MORNING BRIEF
  • INV-04 — INVENTORY VELOCITY
  • REV-05 — REVIEW SENTIMENT
  • OWNER BRIEF — ONE PAGE

FLOOR PLAN // 5 MODULES · 1 LEDGER · 1 BRIEF

Five modules run the night. They read the ledger, watch for signals, and write one brief by morning. Five modules run the night: read the ledger, watch for signals, write one brief. That is the whole method.

▸ READ WHAT THEY WRITE — FREE

▸ READ WHAT THEY WRITE — ONE FULL OWNER BRIEF, FREE

THE SHIFT

5 MODULES / 0 PEOPLE

SYS / ARMED — 0 CALLS

SHOPIFYSETTLED
00:12 · 31 ORD1,240.00
00:31 · 24 ORD980.50
00:58 · 49 ORD1,999.50
TOTAL4,220.00
SHOPEET+2
00:22 · 37 ORD1,430.00
00:47 · 38 ORD1,464.00
PENDING T+2−34.00
TOTAL2,860.00
TIKTOK SHOPTZ −1
00:19 · 17 ORD655.00
00:52 · 18 ORD765.00
TOTAL1,420.00
SHOPIFY · SETTLED4,220.00
SHOPEE · T+22,860.00
PENDING T+2−34.00
TIKTOK SHOP · TZ −11,420.00
GROSS · 214 ORD8,500.00
GROSS 8,500.00 214 ORDERS PENDING T+2 34.00 TIE-OUT Δ 0.4% — EXPLAINED

TIE-OUT Δ 0.4% — EXPLAINED

AND CLOSED: THE MORNING'S MISSING $340 — SHOPEE T+2 BATCH, SETTLED 00:47

THE MORNING'S $340 — FOUND: SHOPEE T+2, SETTLED 00:47

REC-01 01:12

REVENUE RECONCILIATION

Every settlement row, every channel, landed in one ledger in one currency. Tied out by 01:12 — Shopee's T+2 drift explained, not accepted. One ledger, one currency, tied out by 01:12 — T+2 explained, not accepted. The 45-minute morning pass, performed in zero.

INPUT
SHOPIFY · SHOPEE · TIKTOK SHOP
CHECK
ROW-LEVEL TIE-OUT · FX NORMALIZE
OUTPUT
CLEAN LEDGER → BRF-03
LOGGED
01:12 · CLOSED
SIMULATED FEED

Δ 0.4%

03:14 · CPM +212% · 4.9× MAD

ROLLING MEDIAN + MAD — COMPUTED, NOT DRAWN · SET PAUSED 03:14 · PER YOUR THRESHOLD RULEMEDIAN + MAD — COMPUTED, NOT DRAWN · PAUSED PER YOUR RULE

You slept through this. That was the point.

SIG-02 03:14

AD ANOMALY DETECTION

A rolling median and MAD watched spend and CPM per ad set, all night. At 03:14 one set broke the band and paused itself — under the threshold rule you set; the event went to the brief. Ads that should have been paused three days ago run $2K–$8K a month.

The band broke, the set paused itself, the event went to the brief. Caught at 03:14, not on Friday.

INPUT
META · TIKTOK SPEND FEED
CHECK
ROLLING MEDIAN + 3×MAD
OUTPUT
SET PAUSED → BRF-03
LOGGED
03:14 · PAUSED
SIMULATED FEED

+212%

  • SKU-104 · HERO · 38/DAY
  • SKU-121 · 22/DAY
  • SKU-217 · 6/DAY
  • SKU-104 · COVER 11d < LEAD 14d — REORDER TODAYSKU-104 · HERO · COVER 11d < LEAD 14d — REORDER TODAY
  • SKU-217 · SLOW — TRIM NEXT ORDER
INV-04 04:02

INVENTORY VELOCITY

Sell-through per SKU, read against supplier lead time. At 04:02 the hero dropped under its lead. The reorder call went on the brief — before the stockout, not after. A hero dark for a 14-day lead is the full margin on every unit you couldn't ship.

INPUT
SALES VELOCITY · STOCK ON HAND
CHECK
DAYS-OF-COVER VS LEAD
OUTPUT
REORDER FLAG → BRF-03
LOGGED
04:02 · FLAGGED
SIMULATED FEED

11 DAYS

"SIZING RUNS SMALL"

×31 · 7-DAY ACCRETION · → BRF-03

REV-05 04:47

REVIEW SENTIMENT

Every review landed in one stream and sorted overnight. At 04:47 a repeating phrase became a theme; the theme became a line in the brief. The 1-star on your hero product surfaces at 06:00, not on day 3.Surfaced at 06:00, not day 3.

INPUT
REVIEWS · ALL CHANNELS
CHECK
PHRASE CLUSTERING · SENTIMENT DRIFT
OUTPUT
THEME FLAG → BRF-03
LOGGED
04:47 · FLAGGED
SIMULATED FEED

×31

Everything upstream writes one page.

// 05:30

THE BRIEF

LIGHTS OUT // METHODOWNER BRIEF  OP-17

RUN INDEX / 19 Jul 2026 / 06:00 SGT / WINDOW 23:41 → 06:00 · SIMULATED

OPERATIONS NOMINAL.
4 SIGNALS OBSERVED, 4 RESOLVED OVERNIGHT.
1 DECISION WAITING.

STATUS
NOMINAL
MODULES
5/5
SIGNALS
04
ESCALATIONS
00

KEY SIGNALS

  • REC-01 / LEDGER TIED Δ 0.4% — T+2 EXPLAINED //
  • SIG-02 / CPM +212% — SET PAUSED 03:14 //
  • INV-04 / SKU-104 COVER 11d < LEAD 14d //
  • REV-05 / "SIZING RUNS SMALL" ×31 //

DECISIONS

(!) DECISION — REORDER SKU-104 TODAY · LEAD 14 DAYS

LEDGER

GROSS 8,500    NET 8,466    ORDERS 214

GET THE METHOD — $97 30-DAY REFUND

// 05:45

THE RANGE

One method, four amounts of leverage: read it, watch it built, import it working, or point it at your own operation.

BUILT FOR DTC BRANDS DOING $30K–$500K/MO

THE LIGHTS OUT METHOD

THE TOOLKIT + THE GUIDE · SHEETS + PDF YOU OWN, NOT SAAS

$97

SURFACE A — THE TOOLKIT

Per-chapter ROI calculator (Sheets)
Anomaly threshold decision tree
Lead-time / days-of-cover table + iCal audit
Owner Brief · Run Log · Module Card templates
~20 min to your first ledger check

SURFACE B — THE GUIDE

The 5 modules — mechanism, stack, failure modes
~12,800 words · a 90-minute read
The order that works: REC-01 first

ROI calculator · threshold tree · lead-time table + iCal
Owner Brief · Run Log · Module Card templates
The 5 modules — mechanism + failure modes · 90-min read
Sheets + PDF you own, not SaaS · ~20 min to first check

GET THE METHOD — $97 30-DAY REFUND.

30-DAY REFUND · ▸ SEE THE SPECIMEN FIRST

An ops hire runs $6K a month. A fractional COO, $5–10K. The method is $97.An ops hire runs $6K/mo. The method is $97.

SHOPIFY · SHOPEE · TIKTOK SHOP · META — NO N8N, NO CODE AT THIS TIERSHOPIFY · SHOPEE · TIKTOK SHOP · META — NO CODE AT $97

ANNEX // BEFORE YOU BUY

THE SMALL PRINT

What it won't do, and what happens if you want your money back.

A / FIT

THIS IS FOR YOU IF

  • You run a DTC brand at $30K–$500K a month, across more than one channel.
  • Your morning starts by reconciling numbers by hand, and they don't tie.
  • You can give it a weekend. 90 minutes to read, 20 minutes to the first ledger check.
  • You'd rather own the files than rent another dashboard.
  • You'll run one module for a week before wiring the next.

THIS IS NOT FOR YOU IF

  • You're pre-revenue, or selling on one channel. There's nothing to reconcile yet.
  • You want someone else to build it. That's The Night Shift at $997.
  • You came for ad creative or media buying. Nothing here writes a hook.
  • You expect the $97 tier to run itself. It's a method and a set of templates. The wiring is yours.
  • Your stack looks nothing like Shopify, Shopee, TikTok Shop or Meta. The mechanism carries over; the worked examples won't.

B / QUESTIONS

Everything on this page says SIMULATED. Has any of it run for real?

Yes. What's simulated is the data, not the method. Settlement rows, ad spend, stock cover and review text are confidential to the brand they belong to, so every figure on this page is a stand-in: the ledger, the CPM spike, the SKU, the review count.

The mechanism and the thresholds are the ones that ran. The specimen brief is what a real night produces, with the numbers swapped out.

Do I need to be technical?

Not at $97. The toolkit is Sheets and PDFs, and there's no code in any of it. n8n starts at $397, and that tier says so.

My stack isn't Shopify, Shopee or TikTok Shop. Does it still work?

The method is shaped around channels that export settlements, not around those three names. Any channel with a settlement report and a date column runs the same tie-out. The worked examples use what we've run, so you'll be translating column names rather than following along.

How long until something is actually running?

90 minutes to read the guide. About 20 minutes after that to your first ledger check. Then one module a week — reconciliation first, because everything downstream reads from those numbers.

Is this a subscription?

No. One payment, and the files are yours. The Module Library at $397 carries 12 months of updates, then $97 a year if you want to keep receiving them. The workflows you already imported keep running either way.

What's the difference between $97 and $397?

At $97 you get the method and build it yourself. At $397 you import the five workflows, already wired, into your own n8n instance. About 2 hours of setup. Same modules, different amount of work.

What happens after I pay?

An email with one signed link to your vault, keyed to the address you paid with. It opens for 12 months. Everything inside is a download. Sheets and PDFs on your own drive, not a login you have to keep.

C / TERMS

$97 / $197
30 days. Email ops@lightsout.so and it's refunded. You don't have to explain.
$397
14 days, and until you import. Once the workflows are inside your n8n instance they're yours to keep, so the window closes at import.
$997
Refundable in full any time before the Signal Report ships. After it ships it's final. You have the report.

$97

THE LIGHTS OUT METHOD

The toolkit and the guide. Sheets and PDF you own, not SaaS.

GET THE METHOD — $97 30-DAY REFUND

The creed

THE LIGHTS ARE OUT.

THE WORK IS STILL HAPPENING.

YOU CAN SLEEP.

Lights out. So you can sleep.

QUIET OPS / SYSTEMS RUNNING IN THE DARK

LIGHTS OUT // METHOD RUN LOG RL-06

WINDOW 23:41 → 06:00 · SIMULATED RUN

RUN LOG RL-06 · WINDOW 23:41 → 06:00

00:00  SYS    LIGHTS OUT  · RUNNING
01:12  REC-01  TIE-OUT Δ 0.4%  · CLOSED
03:14  SIG-02  CPM +212%  · PAUSED
04:02  INV-04  COVER 11d < 14d  · FLAGGED
04:47  REV-05  THEME ×31  · FLAGGED
05:30  BRF-03  OWNER BRIEF  · SENT

SHIFT 23:41–06:00 · 4 SIGNALS RESOLVED
1 BRIEF WRITTEN · 0 HUMANS WOKEN

SIMULATED RUNCLASS: BLACK / ARCHIVE

GET THE METHOD — $97 30-DAY REFUND.

THE FIELD JOURNAL

SIMULATED RUN · REAL FORMAT