Skip to content
GuideFinance

AI Invoice Automation: The Complete 2026 Guide (From PDF to Paid)

The definitive SMB guide to AI invoice automation: extraction, duplicates, coding drafts, approvals, pricing for BILL, Melio, Ramp, and Dext — without auto-paying blindly.

AI Growthub StaffEditorial TeamPublished Updated August 13, 202623 min read
Independently reviewedEditorial policyFact-checkingLast updated
AI Invoice Automation: The Complete 2026 Guide (From PDF to Paid)

Vendor PDFs do not need a finance transformation program. They need to become draft bills without retyping, duplicates need to surface before anyone pays twice, and approvals need a named owner — not a shared inbox that swallows the trail.

AI invoice automation is that supervised loop: intake → extract → duplicate check → coding draft → human approve → post or pay. It is not “AI pays my bills.” Done well, it cuts hours and catches costly mistakes. Done poorly, it auto-codes a five-figure duplicate or routes bank details into the wrong tool.

This is the definitive 2026 guide for SMB owners, ops leads, freelancers with bookkeepers, agencies, and consultants in the US, Canada, UK, and Australia who process enough invoices to feel the pain — and refuse to auto-pay from unverified AI output.

It sits under AI automation for small business. Firm-scale patterns live in AI workflow for accountants. Keep AP separate from CRM-led collections covered in AI CRM automation.

Table of contents

  1. Quick summary
  2. What is AI invoice automation?
  3. Who should use it
  4. Who should NOT use it
  5. Quick recommendation
  6. Things to consider before choosing
  7. Key features
  8. Best-for table
  9. Pricing in 2026
  10. Pros and cons
  11. Best use cases
  12. Limitations
  13. Comparison tables
  14. Decision matrix
  15. Setup checklist
  16. Six core workflows
  17. AP vs AR
  18. Security rules
  19. ROI math
  20. Four-week rollout
  21. Common mistakes
  22. Alternatives and competitor comparison
  23. FAQ
  24. Final recommendation

Quick summary

If your main pain is…Start hereUpgrade when…
Retyping vendor PDFsDedicated bills inbox + OCR extraction (draft-only)Field accuracy logged on 50 of your invoices
Paying the same bill twiceDuplicate match on vendor + invoice # + amount + dateNear-duplicates flagged for 2 weeks with zero misses
Coding takes foreverAI category / GL suggestions from a coding bibleCorrection rate under ~25% for two weeks
Approvals live in one inboxAmount-based routing to named approversEvery post has an audit trail
Month-end is a scrambleBatch review queue (15–20 drafts at a time)Weekly payment run stays separate from AI
AR cash is slowOverdue reminder drafts onlyHuman edits and sends every reminder

Default bias for SMBs: extract → duplicate check → approve → post. Payment stays a controlled second step.


What is AI invoice automation?

AI invoice automation is a documented path from vendor document to ledger entry — not a chatbot with your bank login.

The spine:

  1. Intake — Invoice arrives by email, portal upload, or scan into one controlled channel.
  2. Classify — System labels invoice vs credit note vs statement vs junk.
  3. Extract — OCR + AI returns vendor, invoice number, dates, tax, totals, and line items when needed.
  4. Duplicate check — Match against open and paid bills; flag near-duplicates.
  5. Coding draft — Suggest GL / category from history and your rules.
  6. Human gate — Named approver confirms or requests missing info.
  7. Post / pay — Bookkeeper posts; payment run stays separate with dual control where required.

AI is strongest at field extraction, duplicate matching, and coding suggestions. It is weakest at payment authorization, vendor identity verification, and fraud judgment.

The same draft-and-approve habit shows up in daily AI workflow for founders and AI productivity for small business.

Four-stage still life suggesting invoice intake, extraction, review checklist, and approval before payment

Who should use AI invoice automation

Use it if most of these are true:

  • You process roughly 20–30+ vendor invoices per month (or fewer, but high pain / high value).
  • Invoices already (or can) land in one inbox or folder.
  • You have a chart of accounts or category list the model can reference.
  • Someone owns approvals — founder, ops, bookkeeper, or external firm.
  • Manual process already risks duplicates or wrong-vendor payments.
  • You can measure minutes per invoice today for a baseline.

Good fits: multi-vendor service businesses, clinics, agencies, restaurants with many suppliers, freelancers who batch bills with a bookkeeper, and consultants managing client AP under clear segregation rules.

Accountants serving multiple entities should start with AI workflow for accountants for firm security and client boundaries.


Who should NOT use AI invoice automation

Skip dedicated AP AI (or delay it) if:

  • You process a handful of simple bills a month and native ledger capture is enough.
  • Invoices still arrive in personal Gmail with no naming convention — fix intake first.
  • Nobody will review drafts — “set and forget” is how double payments happen.
  • You want the tool to authorize bank transfers without a human. That is a policy failure, not a feature gap.
  • Your only plan is pasting invoices into a consumer chat with no retention policy or business agreement.
  • You are mixing AP automation with aggressive AR auto-escalation in one switch. Keep them separate.

If your bottleneck is lead follow-up rather than vendor bills, fix CRM first — AI CRM automation — then return to AP.


Quick recommendation


Things to consider before choosing

  1. Ledger fit — QuickBooks, Xero, NetSuite, and Sage Intacct are not interchangeable. Confirm two-way sync and which plan unlocks it.
  2. Volume vs seats — Per-user AP tools get expensive with many approvers; document-volume tools get expensive with line-item credits.
  3. Approval design — Amount thresholds, project/class coding, and external accountant roles matter more than marketing “AI accuracy” claims.
  4. Payment vs capture — Some products excel at OCR; others excel at paying vendors. You may need both layers — or one that does both.
  5. Fraud surface — Changed bank details require out-of-band verification. AI cannot clear vendor impersonation. Pair this with how to protect your small business from AI phishing.
  6. Data retention — Prefer vendors with business agreements, role-based access, and audit logs over consumer AI chats.
  7. Total cost — Subscription + ACH/check fees + review time. Cheap software with 40% correction rates is not cheap.
  8. Rollout ownership — Name who owns intake, who approves, and who runs weekly payment. Tools do not replace owners.

Key features of strong AI invoice automation

Dedicated intake

One bills@ address, portal upload, or scanner path. Personal-inbox forwarding gets banned after week one. Automation creates one ticket or bill draft per PDF.

OCR + structured extraction

Vendor name, invoice number, invoice date, due date, currency, subtotal, tax, total, line items, PO reference when present. Nulls beat invented values. Side-by-side PDF + fields help reviewers catch hallucinations fast.

Duplicate and near-duplicate detection

Exact match on vendor + invoice number; near-match on same vendor + amount within a date window. Approver marks false positive or blocks payment. One prevented double payment often funds a year of software.

Coding suggestions with a coding bible

Short living doc: recurring vendors, default categories, capital vs expense rules, class/location norms. AI suggests; humans confirm. Ambiguous vendors stay in review.

Approval routing

Thresholds (for example under $500 ops, over $500 founder), vendor-based rules, and multi-step approvals for larger teams. Mobile approve is fine; silent auto-post is not.

Exception handling

Low confidence or missing PO triggers a vendor email draft from an approved mailbox. Human sends. Case stays open until resolved.

Audit trail

Who extracted, who edited, who approved post, who released payment, and when. Required for month-end and fraud review.

Separated payment controls

ACH, check, card, or wire as a second workflow — ideally dual control. AI should not hold the only key to the bank.

For wiring intake across tools without building an ERP, some teams use Zapier/Make/n8n — see n8n vs Zapier vs Make and AI agent vs chatbot vs Zapier for when connectors are enough vs when you need a real AP product.

Colleagues reviewing vendor invoices together during an approval session

Best-for table

ProfileBest starting stackWhy
Solo founder, <20 bills/moNative QBO/Xero capture + manual postLowest cost; AI layer is optional
US SMB, 30–80 bills/mo, QBOMelio Core/Boost or Ramp FreeCapture + pay without enterprise pricing
Growing team needing AP + ARBILL Team ($65/user/mo)Official Invoice Coding Agent + 2-way sync
Spend platform already wantedRamp Free → PlusCards, expenses, and Bill Pay in one place
Bookkeeper / firm, many docsDext PrepareDocument volume + accounting sync focus
Multi-entity / NetSuiteBILL Corporate/Enterprise or Ramp PlusAdvanced sync and controls
Agency / freelancer batching billsNative capture + weekly human reviewAvoid overbuying seats

Pricing in 2026

Pricing below is summarized from official vendor pricing pages as of August 2026. Transaction fees and regional taxes vary — confirm on the vendor site before buying.

BILL (formerly Bill.com)

According to BILL pricing:

PlanList priceNotable AI / AP notes
Essentials$49/user/monthCentral inbox, approvals, payments; CSV-style accounting sync
Team$65/user/monthAutomatic 2-way sync with QBO/Xero; Invoice Coding Agent
Corporate$89/user/monthCustom policies, procurement features, multi-role controls
EnterpriseCustomNetSuite / Sage Intacct / Dynamics-class sync, SSO, dual control

BILL’s product updates state the Invoice Coding Agent is available on Team and above for direct customers, with predictions shown on bill entry until a user approves. Per-transaction fees (ACH, check, card, wires) apply on top of seats.

Ramp Bill Pay

According to Ramp pricing:

PlanList priceBill Pay notes
Free$0/user/monthAutomated invoice OCR, approval workflows, fraud checks, ACH/card/check/wire
Plus$15/user/month + platform feeAuto-coded line items, AI approval recommendations, batch payments, broader ERP
EnterpriseCustomAdvanced controls and coverage

Ramp markets OCR capture on Free and deeper AI coding/agents on Plus. Strong fit when you want AP inside a broader spend platform.

Melio

According to Melio pricing:

PlanList priceNotes
Go$01 user; AI bill capture; 5 free ACH/month
Core$25/mo + $10/extra userQBO/Xero sync, approvals, batch pay; 20 free ACH/month
Boost$55/mo + $10/extra userAdvanced roles, QBO Desktop; 50 free ACH/month
Unlimited$80/moUnlimited users + unlimited free ACH; NetSuite sync listed

After included ACH allotments, Melio lists additional ACH at $0.50 (confirm current fee table).

Dext Prepare

Based on publicly listed business plans, Dext typically prices by users + monthly document allowance (entry business tiers often land around the mid-$20s/month on annual billing for ~5 users / ~250 documents, with line-item and statement extractions as limited credits). Treat Dext as a capture/publish layer into Xero, QuickBooks, or Sage — not always a full payment network.

Native ledger capture

QuickBooks Online Receipts and Xero Hubdoc (included on many Xero plans) cover low-volume header extraction at little or no extra software cost. They are weaker on multi-step approvals, PO matching, and high-variance line-item PDFs.


Pros and cons

Pros

  • Cuts retyping time on repetitive vendor formats
  • Surfaces duplicates before money leaves the account
  • Creates a clearer approval trail than email threads
  • Improves month-end speed when review is batched
  • Scales better than hiring admin hours for pure data entry

Cons

  • Wrong setup can accelerate bad payments
  • Subscription + transaction fees stack quickly
  • OCR still fails on poor scans and unusual layouts
  • Coding drift happens if the coding bible is stale
  • Consumer AI chats are a poor vault for bank details

Best use cases

1. AP intake and classification

Trigger: PDF/image hits bills@ or upload folder.
AI task: Classify invoice vs credit note vs statement vs spam; extract core fields; flag unreadable areas.
Human step: Confirm it is a bill to pay — not a statement for reference only.

2. OCR and structured extraction

Trigger: Document classified as vendor invoice.
AI task: Structured fields + line items when required; null for unreadable values.
Human step: Correct fields in a review UI before creating the draft bill.

3. Duplicate detection

Trigger: Extraction complete.
AI task: Match open/paid bills; flag near-duplicates (same vendor + amount within ~7 days).
Human step: Approver clears false positives or blocks payment.

4. Approval routing and coding drafts

Trigger: Clean extraction, no duplicate block.
AI task: Suggest category/GL from vendor history + coding bible; route by amount.
Human step: Approver confirms code and project/class.

5. AR reminder drafts (collections assist)

Trigger: Invoice overdue past N days in QBO, Xero, Stripe, or your AR tool.
AI task: Draft polite reminder with invoice number, amount, due date, payment link.
Human step: Ops/AR owner edits and sends. No threats, no auto-escalation.
Tone and sequence craft overlap with AI email automation sequences — still keep AR separate from AP queues.

6. Exception and missing-info loops

Trigger: Low confidence or required field null.
AI task: Draft vendor email requesting PO, remit details, or a clean PDF.
Human step: Send from approved mailbox; log until resolved.

Hands sorting paper invoices into trays during batch AP processing

Limitations

  • AI does not verify vendors. Changed bank details need a call to a known number — see also spear phishing risks.
  • Accuracy claims are vendor-averaged. Your messy multi-page PDFs may differ. Log your correction rate.
  • Native capture is not full AP. Approvals, PO matching, and payment controls often need another layer.
  • Auto-post is not a badge of honor. It is a risk setting.
  • AR and AP share PDFs, not policies. Do not share one auto-send switch.
  • International tax and multi-currency still need human judgment on edge cases.
  • Generic chat extraction is fine for sandbox learning, not for production ledger sync and audit trails.

Comparison tables

Table 1 — Tool comparison for SMB invoice automation

CapabilityNative QBO/XeroMelioBILLRampDext
Primary jobCapture into ledgerPay + captureAP/AR platformSpend + Bill PayDoc capture to ledger
AI extractionBasic OCRAI bill captureExtraction + coding agent (Team+)OCR Free; deeper AI on PlusStrong OCR / line items (credits)
ApprovalsLimitedCore+YesYesExpense-oriented
Duplicate focusThinReview-drivenStrong AP controlsFraud checks + matchingDuplicate detection on docs
Payment networkBank feeds / bill pay add-onsYesLarge vendor networkACH/card/check/wirePayments often separate
Best fitLow volumeCost-sensitive SMBGrowing AP/AR teamCard-centric teamsBookkeepers / high doc volume
Watch-outOutgrows youACH allotments / feesPer-user costPlatform fees on PlusCredit limits on line items

Table 2 — Cost drivers (illustrative, not a quote)

Cost driverWhat to checkWhy it matters
SeatsApprovers + accountantsBILL/Ramp Plus scale with users
DocumentsMonthly PDF volume + line-item needsDext credits and overages
PaymentsACH/check/card/wire feesMelio/BILL/Ramp fee tables differ
Sync tierQBO vs NetSuite vs multi-entityOften locked behind higher plans
Review laborMinutes per invoice after AIDominates ROI for 90 days

Decision matrix

Score each option 1–5 for your business. Highest total wins — then run a 50-invoice sandbox before production.

Criterion (weight)Native ledgerMelioBILLRampDext
Fit to your accounting stack (×3)
Approval & audit needs (×3)
Invoice volume & PDF variety (×2)
Payment controls needed (×2)
Total cost predictability (×2)
Time-to-value this month (×1)
Weighted total

Rule of thumb: if approvals and dual control matter more than free ACH, do not pick on sticker price alone.


Setup checklist

Use this before you enable any AI coding or payment feature:

  • Dedicated bills@ (or portal) live; personal inbox forwarding banned
  • Baseline: minutes per invoice on 10 recent bills
  • Chart of accounts / coding bible exported (1–2 pages is enough)
  • Named approvers by amount threshold
  • Duplicate rule defined (exact + near-match window)
  • Tool connected to ledger in draft-only mode
  • Sandbox: 50 historical invoices processed; accuracy logged by vendor
  • Correction-rate target set (pause expansion if >25% material fixes after 2 weeks)
  • Payment run schedule separate from AI extraction
  • Vendor bank-change policy: out-of-band verify only
  • Access limited to finance roles; audit log reviewed weekly
  • Exception mailbox template approved for missing PO / bad PDFs

Six core workflows

Example: ~40 invoices/month SMB

Intake — Vendors send to bills@ only. Automation creates one ticket per PDF.
Extraction — Fields land in a review queue; bookkeeper batches 15–20.
Duplicates — System flags matches; founder catches a recurring SaaS double bill in month one.
Approval — Under $500 ops lead; over $500 founder before post.
Payment — Weekly run is separate; no AI step triggers bank transfer.

Prompt patterns (draft-only experiments)

If you prototype extraction outside an AP product, keep prompts strict and never paste bank credentials:

Extract bookkeeping fields from this invoice for draft entry only.
Return JSON with keys:
vendor_name, invoice_number, invoice_date, due_date, currency,
subtotal, tax_total, total,
line_items (description, qty, unit_price, line_total),
missing_fields (array), low_confidence_fields (array).
Use null for unreadable fields. Do not invent values.
Do not assign final GL codes.
Suggest a GL/category code from the coding bible below.
Return JSON: suggested_code, confidence (high|medium|low), reason.
If confidence is not high, set needs_review true.
Vendor: {{vendor}}
Line summary: {{line_summary}}
Coding bible:
{{coa_rules}}

Move to a product with a review UI once you leave the sandbox.


AP vs AR: keep the workflows separate

AP (bills you owe) prioritizes vendor identity, duplicates, GL coding, and payment controls. Worst failure: paying the wrong vendor twice.

AR (invoices you are owed) prioritizes accurate customer records, polite tone, and correct payment links. Worst failure: damaging a customer with an aggressive auto-email.

Same extraction tools can serve both. Use separate queues and never share one auto-send switch. Label tickets AP or AR at intake so prompts pull the right template.


Security rules for invoice AI

  1. Approved tools only — business agreements and retention policies.
  2. Minimize data — send the invoice PDF, not your full vendor folder.
  3. Never paste bank credentials into any AI chat.
  4. Vendor change verification — call a known number; AI cannot clear fraud.
  5. Access control — finance roles only on the extraction queue.
  6. Audit log — who approved post and payment, when.
  7. Sandbox first — 50 historical invoices before production.

ROI math that finance owners trust

Net hours saved = (manual minutes − AI-assisted minutes including review) × monthly volume − tool and maintenance cost

Example:

  • Manual: 12 min × 40 invoices = 480 min (8 hrs)
  • AI-assisted: 5 min × 40 = 200 min (3.3 hrs)
  • Net: ~4.7 hrs/month before duplicate-prevention value

If more than ~25% of extractions need material fixes after two weeks, improve scan quality, vendor templates, or switch extraction tool — do not add auto-pay.


Four-week rollout plan

WeekFocusExit criteria
1Centralize intake + baseline timingAll invoices to one inbox; 10-invoice time study
2Extraction on 50 sample PDFsField accuracy log by vendor
3Duplicate rules + coding draftsApprover queue live; no auto-post
4Production batch reviewNamed owner, weekly error review, ROI note

Common mistakes

Auto-post on first extraction success

Fix: Draft-only until correction rate is acceptable. Posting requires an approver click.

Treating ChatGPT as your AP system

Fix: Use it only for learning prompts. Production needs sync, roles, and audit logs.

Skipping duplicate detection

Fix: Turn on exact + near-match before you celebrate speed gains.

Mixing AP automation with payment automation too early

Fix: Separate workflows. Keep dual control on payment runs.

Ignoring vendor bank-detail changes

Fix: Out-of-band verification every time. This is a people process, not an OCR setting.

Chart of accounts drift

Fix: Update the coding bible monthly. Ambiguous vendors stay in review.

Buying seats before fixing intake

Fix: One bills@ channel first. Software cannot organize chaos it never receives.


Alternatives and competitor comparison

BILL vs Melio vs Ramp (major competitor lens)

NeedPrefer BILLPrefer MelioPrefer Ramp
Strongest AP/AR platform depthYes — Team/Corporate feature setGood SMB AP/ARAP inside spend platform
Lowest entry costNo — from $49/user/moYes — Go free / Core $25Yes — Free Bill Pay tier
AI multi-line coding agentOfficial Invoice Coding Agent (Team+)AI capture + assistantDeeper auto-coding on Plus
Card + expense bundleSeparate Spend & Expense productPayment-firstNative strength
QBO/Xero SMB simplicityStrong on Team+Very strongStrong on Free/Plus

Practical take: Melio or Ramp Free for cost-sensitive teams paying mostly domestic vendors; BILL when you want a fuller AP/AR control plane and official coding agents; Ramp when cards, expenses, and bill pay should share one policy brain.

Other alternatives

  • Native QBO/Xero (+ Hubdoc) — best “do nothing extra” path under low volume.
  • Dext — best when document capture quality for bookkeepers matters more than vendor payment networks.
  • Zapier/Make/n8n + document AI — flexible but you own the glue, security, and edge cases (n8n vs Zapier vs Make).
  • Accounting firm workflow — outsource the review queue; keep approvals in-house (AI workflow for accountants).

Suggested articles to publish next

  • BILL vs Melio vs Ramp: which AP stack fits a 50-invoice SMB?
  • QuickBooks Online AI and Receipt Capture: what is enough before you buy AP software?

Frequently asked questions

What is the safest first step in AI invoice automation?

Centralize invoice intake and run field extraction in draft-only mode with human review. Do not enable auto-posting or auto-pay until you have logged accuracy on your real invoice formats.

Can AI invoice automation replace my bookkeeper?

No. AI speeds extraction, duplicate detection, and coding suggestions. A human should approve postings, handle exceptions, and own payment runs.

How does duplicate detection work?

After extraction, match vendor name, invoice number, amount, and date against open and paid bills. Flag exact and near-duplicates for approver review before payment.

Is it safe to upload invoices to ChatGPT?

Use approved business tools with clear data policies. Minimize what you upload, never include bank credentials, and prefer dedicated AP products tied to your accounting system.

What ROI should a small business expect?

Many SMBs save several hours per month on data entry alone, plus value from prevented duplicate payments — but only if review time and correction rates are included.

BILL, Melio, or Ramp — which should I choose?

Choose Melio or Ramp Free for lower entry cost and straightforward US SMB bill pay; BILL Team+ when you need deeper AP/AR controls and the Invoice Coding Agent; Ramp Plus when AI coding and spend policies should live with your corporate cards. Confirm sync with your ledger first.

When is native QuickBooks or Xero capture enough?

When volume is low, invoices are simple, and one person can review drafts weekly. Move to a dedicated AP layer when approvals, duplicates, or PDF variety outgrow the native queue.

Should AP and AR share the same automation?

Share extraction technology if you want — not approval queues or auto-send switches. AP failures hurt cash; AR failures hurt relationships.


Final recommendation

Build one loop this month: PDF in → extract → duplicate check → human approve → post. Keep payment as a separate controlled run.

Pick tooling from volume and ledger reality, not from the loudest “touchless” demo. According to official pricing pages, Melio and Ramp offer the cheapest on-ramps; BILL concentrates AP/AR controls and coding agents at Team and above; Dext and native Hubdoc/Receipts cover capture-heavy or low-volume paths.

Expand only after correction rates earn trust. The first prevented duplicate usually pays for the stack — and earns the right to automate the next step.

Key takeaway

The definitive SMB guide to AI invoice automation: extraction, duplicates, coding drafts, approvals, pricing for BILL, Melio, Ramp, and Dext — without auto-paying blindly. For more step-by-step guides, browse our blog or explore Finance.

Frequently asked questions

What is the safest first step in AI invoice automation?

Centralize invoice intake and run field extraction in draft-only mode with human review. Do not enable auto-posting or auto-pay until you have logged accuracy on your real invoice formats.

Can AI invoice automation replace my bookkeeper?

No. AI speeds extraction, duplicate detection, and coding suggestions. A human should approve postings, handle exceptions, and own payment runs.

How does duplicate detection work with AI invoices?

After extraction, match vendor name, invoice number, amount, and date against open and paid bills. Flag exact and near-duplicates for approver review before payment.

Is it safe to upload invoices to ChatGPT?

Use approved business tools with clear data policies. Minimize what you upload, never include bank credentials, and prefer dedicated AP products tied to your accounting system.

What ROI should a small business expect from invoice automation?

Many SMBs save several hours per month on data entry alone, plus value from prevented duplicate payments — but only if review time and correction rates are included.

BILL, Melio, or Ramp — which should I choose?

Choose Melio or Ramp Free for lower entry cost and straightforward US SMB bill pay; BILL Team+ when you need deeper AP/AR controls and the Invoice Coding Agent; Ramp Plus when AI coding and spend policies should live with your corporate cards. Confirm sync with your ledger first.

When is native QuickBooks or Xero capture enough?

When volume is low, invoices are simple, and one person can review drafts weekly. Move to a dedicated AP layer when approvals, duplicates, or PDF variety outgrow the native queue.

Should AP and AR share the same automation?

Share extraction technology if you want — not approval queues or auto-send switches. AP failures hurt cash; AR failures hurt relationships.

Written by

AI Growthub Staff

Editorial Team

The AI Growthub editorial team covers practical AI news, tools, and workflows for small business owners. Every article is fact-checked against primary sources before publication.

Comments are coming soon

We’re building a discussion space for business owners. Until then, reply to any newsletter issue — we read everything.

Free weekly briefing · every Tuesday

The AI edge, delivered every Tuesday

One 5-minute email: the tools worth your money, the plays that are working right now, and zero hype. Unsubscribe anytime.

No spam. No selling your data. Read by owners of restaurants, gyms, clinics, and agencies across the US, UK, Canada, and Australia.