Margin alerts
The Friday margin read, retired.
The weekly margin digest the bookkeeping loop fires the moment the week’s per-day batches settle — per-line margin walk, top-three spend attribution, next-digest scheduled — so the margin read the owner rebuilds from POS tape + processor report + bank ledger stops being the Sunday the operator writes it by hand. The full wire-once beat is on /how-it-works; the Starter tier is where the weekly margin digest ships first.
The Sunday-morning margin read.
There is another tax small operators pay — one the bookkeeping books won’t show on a trial balance, but the operator’s Sunday-morning hours carry all the same. It is paid in three open tabs and a hand-typed spreadsheet: the POS tape on the left monitor, the payments processor report on the right, the bank ledger in the second browser window — the operator hand-subtracts the cost-of-goods from the receipts, ignores the tip batch that left against the till and not the batch, and writes a margin number onto an emailed-to-self spreadsheet that will go unopened the whole week.
The number is right about eighty percent of the time. The other twenty percent it’s wrong by enough that the subscription the operator didn’t remember renewing shows up as a top-three spend a week late — the line item the spreadsheet never had a bucket for, the cost-of-goods the POS posted the day after the spreadsheet saved, the three-week-old labor line the operator forgot to roll into the read. The margin lands one way at eight and a different way at ten, and the operator finds out which one was right the first time a renewal auto-debit lands without the cushion to cover it.
By the next Sunday the operator is reconciling three feeds against the emailed-to-self spreadsheet — the POS tape, the processor report, the bank ledger — and re-checking each against the per-day batch the POS posted last week, in case a comp or a void-out moved a margin between the spreadsheet save and now. The hours are paid in an unscheduled Sunday, a missed family breakfast, the quiet panic the line-of-credit statement carries when it lands on a Wednesday morning with a margin the operator would have caught on Friday if the read had been right.
That tax — the one three feeds publish a margin for, the one the bookkeeping books already carry as a line item in the period close, the one the operator rebuilds by hand on the Sunday before the books go out — is the one Stillpost built the weekly margin digest to retire. Same three feeds. Same three sources. A read-only connect per source, one wire-up. The per-line margin is folded against the cost-of-goods the operator picked during wire-up, the top-three spend categories are attributed against the week-before baseline, and the read the operator opens on Monday morning was last computed Friday evening, before the till opened for the first rush. The Sunday before the books go out is the Sunday your CPA closed last Friday, not the one you will be writing into this morning.
That is the beat. That is the loop. That is what gets handed back.
What the digest does on a Friday pull.
Three beats the weekly margin digest runs after the wire-up — in the order the loop fires them. Tally, attribute, schedule. Each beat folds into the next, and the per-line margin the operator opens on Monday is the per-line margin the loop posted Friday evening, not the one the operator wrote onto an emailed-to-self spreadsheet an hour after the till closed.
- Step one01/03
Tally the per-line margin.
Every receipt the bank ledger accepted that week gets folded against the cost-of-goods the POS tape line posted — discount, comp, void-out, the tip that left against the batch not the till — and the per-line margin lands on the same row the seller’s invoice already carries, so the loop can read the difference without an analyst between.
- Step two02/03
Attribute the top three spend categories.
The week’s outflow rolls up against the buckets the operator picked during wire-up — COGS, labor, the four-letter subscriptions that quietly drain a till-month — and the digest calls out the three that moved the most against the week-before baseline. No twelve-row spreadsheet the operator reads on a Sunday, no mention of a category the operator doesn’t run, just the three that need a decision this week.
- Step three03/03
Schedule the next digest.
The next Friday lands in the calendar before this one closes — the loop fires again once the per-day POS batch settles and the bank ledger has cleared the prior week — and the operator walks into Monday morning with last week’s read already in the inbox, before the till opens for the first rush.
What it earns its keep on.
Three returns the weekly margin digest ships, measured the way an operator measures them — the Friday read retired, a margin the till opens to, three buckets not twelve.
The Friday read, retired.
The Sunday-morning ritual of opening the POS tape, the processor report, and the bank ledger to hand-subtract cost-of-goods from receipts is a ritual the loop retired. The Sunday that used to build the number is the Sunday the operator got back.
A read the till opens to.
The digest lands Friday evening — after the per-day POS batch settles, after the bank ledger has cleared the prior week, before the owner opens for the first rush Monday — so the first decision of the week is informed by last week’s read, not by the sticky note on the laptop hinge.
Three buckets, not twelve.
The loop attributes the top three spend categories the operator picked during wire-up, against the week-before baseline, on a rolling Friday cadence. The receipt-by-receipt spreadsheet the operator used to email themselves on Sunday is the spreadsheet nobody opens anymore.
The five questions operators ask before they wire it up.
What the digest computes, when the Friday pull fires, whether a service business without a till sees the same read, what a per-day batch not-yet-settled looks like, and how the spend buckets tune — straight answers, before the wire-once walk starts. The FAQPage-structured-data block above carries the same questions in a machine-readable shape Google reads directly.
What does the weekly margin digest actually compute?
Per-line margin for every receipt the bank accepted that week. The loop reads the per-line receipt the POS tape already posted — discount, comp, void-out, the tip that left against the batch and not against the till — folds it against the cost-of-goods the operator wired up during onboarding, and posts the per-line margin on the same row the seller’s invoice already carries, so a line the operator opens in the digest backs out to the line the buyer paid against. The week’s outflow rolls up against the three buckets the operator picked during wire-up — COGS, labor, the four-letter subscriptions that quietly drain a till-month — and the digest ranks them against the week-before baseline, calling out the one that moved the most. No twelve-row spreadsheet the operator reads on a Sunday, no mention of a category the operator doesn’t run, just the three that need a decision this week. The per-line row and the per-category roll-up are visible from the dashboard; the Friday email links straight at them.
When exactly does the Friday digest fire?
Friday evening, after the per-day POS batch settles and the bank ledger has cleared the prior week — typically 21:30 local time on the Friday that closes a Sun–Sat work week, occasionally later when a card network batch is still in flight. The digest lands in the operator’s inbox before the weekend starts and is the read the operator walks into Monday morning with, before the till opens for the first rush. The cadence isn’t a poll-every-minute feed — it’s a deliberate answer to the question "how fresh does this read have to be for the operator to act on it without rebuilding it on Sunday?" The window the loop waits for is the same window the CPA waits for, so the per-line margin the digest posts is the per-line margin the books will carry on Monday.
Does it work for a service business, not a till-and-counter shop?
Yes. The weekly margin digest reads the same bank + inbox + point-of-sale feeds the cash-flow watch reads — and folds them against the per-invoice margin the operator wires up during onboarding. A service business picks the per-invoice line the invoice itself carries, a product business picks the per-line COGS the POS tape already posts, and the digest renders the same Friday-evening read either way. The one read that doesn’t apply is the till-count — a service business without a till doesn’t need one — and the loop detects this at wire-up rather than carrying a half-applicable read into the digest. The contact page routes a non-fitting operating model to the team within one business day; the rest of the bookkeeping keeps running in the meantime.
What happens when the per-day POS batch doesn’t settle by Friday evening?
The digest is held. The loop waits for the missing batch, logs the gap as it logs any other feed-down, and fires the digest the moment the batch lands — typically the same weekend, occasionally the Monday morning before the till opens. The per-line margin the digest posts is the per-line margin the batch actually settled on, not the per-line margin the loop guessed at to stay on schedule. When a batch is missing for more than twenty-four hours, the digest surfaces the gap explicitly — the operator opens the email and sees which day’s batch was held — and the contact page routes a stuck batch to the team within one business day. The audit trail the CPA opens on Monday morning shows which day was held, which day settled, and what the per-line margin computation used as the cost-of-goods the loop eventually read.
Can I tune the spend categories the loop attributes?
Yes. The operator picks the three (or five, when the operator wires up a fuller set) buckets during the wire-up walk — COGS, labor, the seven four-letter subscriptions that quietly drain a till-month, whatever buckets the operator actually runs decisions against — and the loop holds them stable across digests so the week-over-week delta is a like-for-like read, not a re-bucketed baseline. Rebucketing is a dashboard thing — the operator changes a bucket, the loop picks it up on the next digest, and the week-before baseline is recomputed against the new bucket retroactively so the very next Friday read is a same-bucket comparison. No analyst gating the rebucket; the operator composes their own category set, and the digest walks from the set the operator chose.
5 questions answered · the loop that fires the digest is on the small-biz pillar, the price frame is on the pricing grid
The next step
Wire the digest today, retire the Friday read.
Starter (under $81 / month) ships the weekly margin digest on top of the bank + inbox + point-of-sale wire-up — the per-line margin walk, the top-three spend attribution, the next-digest scheduled before Monday opens the till. Pro layers the multi-state sales-tax estimate on top. Managed hands the period close to a human. Three tiers, no surprise implementation fee — pick the one that fits your ops.
Wire it up today.
Cancel any month · switch tiers at the next renewal · reply within one business day on any question.
Pick a tier →See the full approach →not ready to pick a tier? Send a note →
reply expected within one business day