Does Toast have a moat?
Toast has a real moat, built on payments volume and daily embedding in the restaurant, but the software layer on top of it is copyable and the lock-in is partly contractual.
Analysis based on public information as of November 19, 2026. Updated November 19, 2026. Not investment advice.
75Castle
Moat Score out of 100. Durable, compounding advantages. Keep digging.

Who could copy Toast?
Non-technical person with AI
Impractical to copy, unlikely to try
A non-technical person with AI cannot copy Toast in any way a restaurant would switch to; they can only copy thin slices like a menu site, QR ordering or marketing helpers, and those slices don't touch the payments, hardware and contract lock-in that hold the installed base.
Developer with an AI coding agent
Years to copy, unlikely to try
A solo developer with a coding agent can copy Toast's software in weeks, but nobody can copy the payments rails, hardware base and installed restaurants, so a credible copy that wins real Toast customers is not realistic.
Funded AI-native startup
Years to copy, likely to try
A funded AI-native team can build a credible software layer in 9-18 months, but copying Toast's whole business (payments, hardware, installed base) takes years and tens of millions of dollars, so the realistic attack is a wedge, not a clone.
AI lab or platform
Years to copy, unlikely to try
An AI lab or platform is unlikely to absorb Toast whole within 18 months, but it can take the ordering, voice and marketing features around the edges while Toast keeps payments, hardware and the in-store system of record.
Your own customer, in-house
Years to copy, unlikely to try
A restaurant could not replace Toast in-house within 30 days, because the hard parts are payments, hardware and compliance, not software; an operator could only copy small slices such as reporting, scheduling or marketing.
What matters most
Biggest threats
- 1Assistants and agent protocols routing diner orders around Toast, turning it into a back-end
- 2Square and other payments-backed rivals matching Toast's AI features on similar economics
- 3AI-native wedges rebuilding subscription workflows and targeting contract expirations
Strongest defenses
- 1In the money flow: $195.1B gross payment volume as a registered payfac
- 2Primary daily workspace with workflow, hardware and contract embedding
- 3Gated operational data and partial system of record for orders and menus
Does Toast have a moat?
Toast has a real moat, and it sits in the money flow, not the software. $195.1B of payment volume, payfac registration and daily use as the restaurant's primary workspace make it hard to displace. The weakness is everything above that. Coding agents can rebuild the subscription workflows and the new AI agents. Square already ships an agentic voice kiosk. Retention leans on contracts and proprietary hardware that operators resent, and shrinking hardware revenue suggests that lock-in is fading. Diner ordering may also move into assistants like ChatGPT, which puts someone between Toast and the order. The single most important move is to become the system agents must write to: ship a merchant-consented, agent-facing ordering and menu API before someone else's protocol turns Toast into a back-end.
What protects Toast?
Proprietary data (moderate): Order, menu, sales and labor data from live restaurants isn't public, and API access is gated by partner tier and merchant consent (doc.toasttab.com). Toast's rights to use that data for training are unconfirmed, because the API terms were only seen as a snippet.
Data effects (weak): Toast IQ gives upsell recommendations from each operator's own sales data (Spring 2026 release). I found no evidence that pooled cross-merchant learning improves the product in a way that compounds.
Regulatory and compliance standing (moderate): Toast is a registered Visa/Mastercard payment facilitator and carries network-rule and fraud obligations (10-K). It relies on third-party processors, and I found no money-transmitter licences. Rivals can rent similar rails, so this is a real but not years-long barrier.
Transaction embedding (strong): Toast processed $195.1B of gross payment volume in FY2025, and fintech was about 82% of $6,153M revenue (10-K, my arithmetic). Square and Clover are in the same position, and Toast depends on a few processors, so the moat is real but not unique.
System of record (moderate): Toast runs orders, menus, payments and labor, and ordering partners must pull menus through its gated API. I couldn't confirm that customers hold no parallel copy, or how cleanly they can export.
Workflow embedding and switching costs (moderate): The platform covers front of house, kitchen, payments and payroll. Reports cite 2-3 year contracts, termination fees and $5k-$15k of proprietary hardware, but these come from competitor blogs. No churn data exists, and switching guides call leaving doable.
Scale economies (moderate): Toast has $6.2B revenue, $195B in payment volume and a large field sales force. I found no gross-profit split or unit-cost evidence that volume converts into a cost or product advantage.
Human relationship, trust and physical delivery (weak): Toast has field sales, hardware installation and support, which is physical delivery. But 1,402 Trustpilot reviews average 3.1/5 and 235 BBB complaints are cited (aggregator), so trust isn't clearly a strength.
Open ecosystem (weak): Toast has a tiered partner API and a loyalty integration API. I found no evidence that third parties earn meaningful revenue building on it, and there's no marketplace data.
What won't hold?
Encoded business logic or workflows: Subscription software ($936M) is mostly encoded restaurant workflows such as modifiers, tips, KDS and tax. Coding agents can rebuild these, so don't count on them.
Feature set: Toast is shipping AI agents fast (Marketing, Reviews, Offers, Scheduling, SEO), but Square shipped an AI voice kiosk on Oct 14, 2026 and features will be copied within months.
Bundling: The pitch is one system for POS, payments, payroll, hardware and marketing. Restaurants are conservative, so unbundling by agents is slower here, but it is a lever, not a moat.
Integrations count: Delivery, accounting, loyalty and payroll integrations matter to multi-unit operators. Agents make each one cheaper to rebuild, though partner approvals still slow rivals.
How long will the moat last?
In 6 months: about 71 (Moat). Square's agentic voice kiosk (launched Oct 2026) and fast AI release cycles erode Toast's AI-agent features as a differentiator. The payments and workflow core is unchanged.
In 18 months: about 67 (Moat). AI-native wedges rebuild ordering, scheduling and marketing workflows and target customers whose 2–3 year contracts are expiring. Hardware lock-in keeps fading, in line with declining hardware revenue.
In 36 months: about 62 (Moat). Assistant-led ordering (following ChatGPT's DoorDash, Uber Eats and OpenTable integrations) and possible agent protocols put a layer between Toast and the diner. This pressures software margins and the system-of-record role, though Toast likely stays in the Moat band because of payments volume.
What should Toast do?
Transaction embedding (money flow): This is the strongest defense you have. $195.1B of FY2025 gross payment volume and about 82% of revenue run through payments. Rivals can rent similar rails, though, and your 10-K says you depend on a limited number of processors. I could not see gross profit by stream, so I don't know how much margin sits behind the volume. First, pull gross profit for payments versus subscription and see how much of the margin you control. Second, review processor and sponsor-bank terms for the next renewal cycle, and model what direct processing or a second processor would save. Third, add money-movement products that keep cash inside Toast: faster payouts, tip and payroll flows, and financing repaid from sales. Offer them where you already do or plan to. Each one makes leaving more than a software swap.
System of record, with an agent write path: You hold orders, menus, payments and labor, and your API is gated by partner tier and merchant consent. That gating protects you today. It also leaves a gap. If assistants and agent protocols define how orders are placed, someone else's standard makes Toast a back-end. I found no MCP or agent-facing API from Toast. Ship a merchant-consented, agent-facing API for menus, carts, orders and order status, such as an MCP-style server. Keep Toast as the only place orders are confirmed, paid and sent to the kitchen. Require partners to certify. Give merchants a permission screen showing which agents can read or write what. Confirm your API terms on data ownership before launch, because I only saw a snippet of Section 11.
Workflow embedding, earned instead of contractual: Retention today leans on 2–3 year contracts, termination fees and proprietary hardware. Those claims come from competitor blogs, so they are directional. Hardware revenue is down 4% over nine months, and the reviews (3.1/5 across 1,402 reviews, 'trapped' complaints) show operators resent the lock-in. Contract lock-in erodes at each renewal. Show each operator a monthly Toast value report: labor hours saved, upsell revenue from Toast IQ, payments and chargebacks handled, and reports they no longer build by hand. Wire the Scheduling, Reviews and Offers agents into the daily kitchen and floor workflow so removing Toast breaks routines the staff actually use. Start a renewal program for accounts expiring within six months, led by the report, not a discount.
Data effects from pooled, consented benchmarks: Toast IQ upsell and scheduling suggestions use each operator's own sales data. I found no evidence that learning across merchants compounds today. A new restaurant on Toast could start with benchmarks and models trained on thousands of comparable venues. A startup with a handful of locations cannot offer that, and that is a data advantage it can't scrape. First, confirm what your terms let you do with order, menu and labor data, and add clear opt-in where they fall short. Then build anonymized benchmarks by cuisine, region and size: labor percent, menu price tests, modifier attach rates and upsell conversion. Feed them back into Toast IQ and into new-merchant setup. Measure whether recommendations improve as the pool grows. If they don't, drop this bet.
Open ecosystem for agents and voice ordering: Voice and ordering AI is arriving from several directions: Square's agentic kiosk, OpenAI-based drive-thru pilots and assistant-led ordering. You can't out-build all of them. If third parties build on Toast and earn money doing it, each one adds to the reasons to stay on Toast. Today the API is tiered and gated, and I found no evidence partners earn meaningful revenue from it. Create a certified agent partner tier with clear API terms, sandbox access and a listing in the Toast merchant dashboard. Offer revenue sharing to voice-ordering, phone-ordering and demand-channel partners whose orders flow through Toast payments. Publish partner earnings or order volume to attract more builders. Hold the right to review partners that touch customer data.
Human trust through support and physical delivery: Toast has field sales, installation and support, which a software wedge can't copy quickly. But 3.1/5 on 1,402 Trustpilot reviews and 235 BBB complaints (from an aggregator) say this is not a strength yet. Restaurants are conservative buyers, and a POS that fails on a Friday night loses the account. Dependable service is the one thing a rival with a good demo can't fake. Set a published service-hour response target for outages. Staff a dedicated line for dinner-rush failures. Track first-year onboarding quality by installer. Tie support and installer incentives to 90-day and 12-month retention, not just go-live. Publish uptime. Read the 'trapped' and fee complaints in the reviews for themes and fix the top three.
Questions people ask
- Does Toast have a moat?
- Toast has a real moat, built on payments volume and daily embedding in the restaurant, but the software layer on top of it is copyable and the lock-in is partly contractual.
- Could OpenAI, Google or Microsoft replace Toast?
- An AI lab or platform is unlikely to absorb Toast whole within 18 months, but it can take the ordering, voice and marketing features around the edges while Toast keeps payments, hardware and the in-store system of record.
- What's the strongest defense?
- In the money flow: $195.1B gross payment volume as a registered payfac
Sources
- pos.toasttab.com: https://pos.toasttab.com
- pos.toasttab.com: https://pos.toasttab.com/pricing
- pos.toasttab.com: https://pos.toasttab.com/sitemap.xml
- startupowl.com: https://startupowl.com/reviews/toast
- pos.toasttab.com: https://pos.toasttab.com/robots.txt
- pos.toasttab.com: https://pos.toasttab.com/llms.txt
- doc.toasttab.com: https://doc.toasttab.com/doc/devguide/apiOverview.html
- fourweekmba.com: https://fourweekmba.com/how-does-toast-make-money-toast-business-model-in-a-nutshell/
- content.edgar-online.com: https://content.edgar-online.com/ExternalLink/EDGAR/0001650164-26-000057.html?dest=tost-20251231_htm&hash=8eab18e48e15871263f1b1d634df334cbc110b256168a08652783e545edc7b56
- businesswire.com: https://www.businesswire.com/news/home/20260212058106/en/Toast-Announces-Fourth-Quarter-and-Full-Year-2025-Financial-Results
- sec.gov: https://www.sec.gov/Archives/edgar/data/0001650164/000165016426000216/tost-20260930.htm
- sec.gov: https://www.sec.gov/Archives/edgar/data/0001650164/000165016426000162/tost-20260630xexhibit991.htm
- sec.gov: https://www.sec.gov/Archives/edgar/data/0001650164/000165016426000113/tost-20260331xexhibit991.htm
- doc.toasttab.com: https://doc.toasttab.com/doc/devguide/apiComparingMenusAPIV2AndV3.html
- api-updates.toasttab.com: https://api-updates.toasttab.com/announcements/menus-api-v3-filters-menu-entities-for-ordering-partners
- pos.toasttab.com: https://pos.toasttab.com/api-terms-of-use
- squareup.com: https://squareup.com/us/en/press/square-kiosk-self-service-reimagined
- fastcasual.com: https://www.fastcasual.com/news/square-launches-ai-voice-ordering-kiosk-for-restaurants/
- restaurantbusinessonline.com: https://www.restaurantbusinessonline.com/technology/restaurant-delivery-coming-chatgpt
- thehospitalityhangout.com: https://www.thehospitalityhangout.com/blog/ai-drive-thru-ordering-in-2026-what-restaurant-operators-need-to-know-now/
Does your business have a moat?
Five AI attackers, live web research and a published rubric. Free, about five minutes, no account. Or talk it through with Matt on a strategy call.