· Aman Narang

Aman Narang: What It Takes to Build Software for 171,000+ Restaurants — Toast

Vertical-SaaS depth — leaning into operational complexity one category at a time — is a more durable moat than horizontal breadth; payments-as-wedge funds the platform, the long apprenticeship into the CEO seat is a feature not a failure, and under-digitized categories like restaurants hide yield-management opportunity the size of the missing business model.

vertical-saasfounder-ceo-transitionpayments-as-wedgesmbcategory-creationscalingrestaurant-tech0% confidence

Why this is in the corpus

Operator-CEO interview with the co-founder of a ~$20B vertical-SaaS company (Toast) covering the full arc: 9-month no-traction start, the pivot from a checkout app to a full restaurant OS, payments/Android/cloud as the model unlock, hiring an outside CEO for 8 years then taking the seat back, and a clear-eyed thesis on TAM expansion and AI-era yield optimization. Dense with replicable plays and named failure modes.

Summary for skimmers

Toast co-founder Aman Narang on building vertical SaaS for restaurants: niche-first PMF, payments-as-wedge, leaning into complexity as moat, the outside-CEO apprenticeship, and the yield-management opportunity in an under-digitized category.

Briefing

What survives the editorial filter

This page should feel like a smart colleague already listened for you and left only the operating logic worth keeping. Not everything said in the episode makes it through.

Trust signal

Direct episode extraction

Best used for

Decision-grade retrieval metadata not yet added for this episode.

Hold lightly

No explicit downgrade reason stored yet for this episode.

Principles

Durable claims that survive beyond the speaker's biography — each with explicit limits, transferability judgment, and evidence.

Principle

Let customers lead product strategy once you have PMF

After PMF, expand the platform along lines your existing customers are already pulling you toward.

Toast went POS to payments to capital to online ordering/gift/loyalty to payroll/scheduling to back-office inventory, much of it customer-driven. Capital was added because customers wouldn't switch off legacy without it.

Customers lead the roadmap once you have fit; expansion pulled by demand beats expansion pushed by vision.

Principle

Use payments and embedded financial products to make the business model VC-fundable

Embed payments to make hardware/service cheap and the SMB business model fundable.

Toast made hardware and service far cheaper than legacy POS (a new restaurant could get running for under $5K vs $50K) by embedding payments, which carried the economics. Capital products followed because legacy providers had them and customers wouldn't switch without one.

Payments-as-wedge: embed the money rail to subsidize cheap software and fund growth.

Principle

A good manager's first job is to be a good coach

The number-one job of a manager is to be a coach who can do the work with the team.

Because Toast hires many people out of the restaurant industry, it must teach them what it means to work in a corporate environment and be an effective manager — and Narang insists the foundation is coaching, not box-checking.

A manager who can't coach has no credibility; coaching is the foundation everything else rests on.

Principle

Complementary, long-trusted co-founders beat any specific founding-team size

What matters in a founding team is complementarity and prior trust, not whether it's two or three people.

Narang's trio (product/GTM, innovation/technical, CTO) had known each other six or seven years; he extends the "founder" label to an early core group including an HBS intern and Steve's brother — the first key engineering hire.

Complementary skills plus prior trust matter more than founding-team size.

Principle

Go vertical and lean into complexity instead of staying horizontal

Depth on one vertical's operational complexity is a more durable moat than horizontal breadth.

Toast chose to nail every restaurant type rather than focus one narrow segment, betting the overlap plus the edge cases (e.g. powering a bar's pool tables) would compound into a defensible platform. They later carried the same "lean into complexity" thesis into retail (grocery, gas stations).

Vertical depth beats horizontal breadth — complexity is the moat, not the obstacle.

Principle

Hire a manager who has scaled before to build the systems you have never built

Bring in someone who has scaled before to install org systems founders have never built.

Toast brought in Chris (who had scaled CS at Endeca) as CEO in 2015; founders became presidents. It was the closest the company came to not working, and the outside operator's "superpower" was bringing the team together and making the mission/values real.

Hire a leader who has scaled before — they install the systems first-time founders have never built.

Principle

Frugality is a strategic capability, not just thrift

Extreme early frugality buys the runway to reach the proof points that unlock the next dollar.

Toast took no salaries, leaned into equity over cash, and stretched a $500K seed (released $100K at a time) because they didn't know if they'd ever raise more. Narang ties it to a frugal upbringing — "doing more with less is part of my DNA."

Frugality is a strategic capability: weight comp to equity, conserve cash, make the seed count.

Principle

The pivot is what survives — the original product is often just the probe

A failed first product is a probe; the durable business is the pain it reveals in customer conversations.

Nine months of no traction on the phone-checkout app led to the realization that restaurants ran on disconnected 1990s systems with no APIs. Pivoting the conversation from checkout to "how do you like your technology broadly" surfaced the platform opportunity.

The probe fails, the conversation it buys you reveals the durable business.

Principle

Prioritize usage and revenue growth over price extraction early

Optimize for usage and traction before price; cheap pricing buys the learning that compounds.

Narang argued pricing was a distraction early — usage was fundamental to learning, and traction was the purest way to recruit talent and raise capital when you have no credibility. Revenue growth is "one of the more purest ways" to punch above your weight.

Don't optimize price before you have usage — traction is the currency that buys talent and capital.

Principle

Treat momentum as oxygen — manufacture it on every axis

Momentum on any axis — revenue, hiring, engineering, funding — is the oxygen that keeps a fragile startup alive.

The host framed it and Narang agreed: companies are so fragile that constant movement is how to be anti-fragile. Momentum is a "deposit" into the people there — proof something works that gets you to the next mile marker of validation.

Momentum is oxygen — manufacture validation milestones on revenue, hiring, engineering, and funding.

Principle

Give teams guardrails and autonomy instead of approval gates

Set guardrails and grant autonomy; approval gates that approve 99.9% of cases only add latency.

Toast built approval rules to get disciplined while scaling, then found 99.9% of approvals went through anyway — front-line managers approved everything. The lesson: autonomy within guardrails preserves the speed that matters most in SMB GTM.

If your approval gate approves 99.9% of the time, it's latency, not control — use guardrails instead.

Frameworks

Reusable systems and operating models — including when they help and when they break.

Framework

New Ventures: an internal special-ops unit to re-inject zero-to-one DNA at scale

Run a structurally isolated internal incubator with its own comp plans to restore zero-to-one capability at scale.

Toast launched New Ventures (~2021), echoing the Endeca "special operations" team it was born from. Isolated teams with focused comp plans shipped Toast Retail, guest-facing reservations/waitlist, catering/events, and an advertising product, with a high hit rate. The key insight: scale operators and entrepreneurs are different people doing different work.

Restore zero-to-one capability at scale via a structurally separate incubator staffed by entrepreneurs, not scale operators.

Framework

Three-vector location-growth model: core/up-market, international, and adjacent vertical

Grow locations along three vectors — core/enterprise, international, and an adjacent vertical — that intersect.

Toast frames growth as core SMB restaurants + up-market enterprise (Marriott, Applebee's, Denny's), international (London, Sydney, Toronto, Australia), and retail (grocery, gas, convenience). Retail itself follows the same SMB-first-then-up-market-then-global pattern, and the vectors compound at intersections.

Map location growth onto three intersecting vectors: deepen core + up-market, go international, replicate into an adjacent vertical.

Framework

Two server questions to read a restaurant's POS in the wild

Ask any restaurant's server two questions — how they like the product and the service — to read the market.

Narang turned this into a personal ritual: in any restaurant he doesn't know, he asks the server (or manager) the two questions restaurateurs care about most. It keeps him connected to real sentiment and competitor performance even as CEO of a 6,500-person company.

Build a repeatable field-research ritual: ask front-line users the one or two dimensions they weigh most.

Framework

Three pillars that made the SMB SaaS model work: cloud, Android hardware, embedded payments

Cloud + commodity Android hardware + embedded payments is the three-pillar recipe for fundable SMB vertical SaaS.

Toast's diagnostic for why it could win where legacy POS couldn't: cloud beat on-prem basement servers, Android handhelds let a new restaurant start under $5K vs legacy's $50K, and embedded payments carried the unit economics so software/hardware could be cheap.

Diagnostic for SMB vertical SaaS: are cloud, cheap hardware, and embedded payments all present? Missing one breaks it.

Signals

What appears to be shifting, for whom it matters, and what happens if you ignore it.

Signal

Conversational data interface (Toast IQ) is the fastest-adopting product because it makes the platform accessible

A ChatGPT-style natural-language layer over proprietary data is the fastest-adopting product because it makes the platform accessible.

Toast IQ lets restaurateurs query all their Toast data, make back-end changes (e.g. "86 an item," revamp a menu), and self-serve support in natural language. It is among Toast's fastest-adopted products and already handles ~half of support volume (the support agent was bought/partnered, not built).

Expect natural-language interfaces over proprietary vertical data to outpace feature products by making the platform accessible.

Signal

Restaurant yield management is the next greenfield as AI lowers the cost of demand-supply matching

AI makes per-restaurant yield optimization feasible for the first time — a greenfield in an under-digitized category.

Narang frames yield as a long-standing gap: restaurants match demand and supply only in analog ways (happy hour). With AI generating offers from demand patterns, weather, and local events, an automated yield engine can fill open slots at ~80% incremental margin — feasible now because AI lowers the matching cost the fragmented market never paid for.

Watch under-digitized fragmented categories with airline-like fixed-cost economics — AI makes yield optimization newly buildable.

Opportunities

Only included where there is a buyer, a real wedge, and a plausible revenue path — not vague idea theater.

Opportunity

Toast Retail: local-commerce TAM 10x larger than restaurants

Local commerce ($7-8T non-e-commerce in-store spend) is a 10x+ TAM expansion for a restaurant vertical platform.

Toast's team sized it: ~$1T restaurants vs $7-8T of relevant local commerce (grocery, gas, convenience, clothing) excluding regulated categories (pharmacies, hospitals). COVID pull (restaurants doing retail to survive) seeded it; Narang believes retail could eventually be bigger than restaurants.

An under-digitized adjacent category sharing the same complexity profile can 10x a vertical platform's TAM.

Opportunity

The personalized, walk-out, pre-fired restaurant experience ("Cheers experience")

Connecting the guest profile to the POS unlocks personalization, pre-firing, and walk-out checkout — a consumer layer on the B2B base.

Narang's vision: book through Toast Tables with card on file, walk out like leaving a hotel/cab, have the kitchen pre-fire your kids' order on check-in, and get a "Netflix-knows-what-you-want" / "Cheers everybody-knows-your-name" personalized recommendation. The wedge is the guest profile living in the POS where restaurants are normally blind to who the guest is.

Where a vertical platform already holds the transaction data, the guest-personalization layer (recommend, pre-fire, walk-out) is the next opportunity.

Lessons still worth keeping

Useful takeaways that did not fully clear the bar for durable principle status.

Lesson

Scaling before the foundations were ready forced a deliberate slowdown

Toast hit early customer pull but had to deliberately slow down to fix offline software, hardware, and support before re-accelerating.

Despite strong demand (cloud, mobile handhelds, sub-$5K setup), the software didn't run offline (fatal when a restaurant's internet drops at 9pm), ~100% of first-gen hardware came back within ~180 days, and support was a Google Voice number that rang the whole company. Co-founders and investor Steve Papa insisted on slowing down.

Demand pull is not readiness — scaling on broken foundations destroys the experience that created the pull.

Lesson

The first product (phone checkout) got nine months of zero traction

Toast's phone-checkout app got nine months of zero traction because legacy systems had no APIs to connect to.

The biggest blocker was structural — there was no way to scale a checkout app on top of disconnected legacy POS with no API connectivity. The failure surfaced the real opportunity (replace the whole stack) during customer conversations.

A structurally unscalable first product (no integration surface) still earns the customer conversations that reveal the real business.

Lesson

Network-only early hiring hid a distribution-of-performance problem at scale

Toast's first 20 hires from its network were all good, but fast hiring without a system produced a huge performance spread.

Narang names this a blind spot: the founder-interviews-everyone proverb didn't get executed well as they scaled past their network in 2014. The huge performance differences were the tell that they lacked the hiring, assessment, and onboarding infrastructure — which they later built with the incoming CEO.

Past your network, fast hiring without an assessment system produces a performance spread that exposes missing org foundations.

The Plays

Try these this week

Verb-first executable actions — each one tied to a stated outcome in the episode.

Make founder disagreements invisible to the team by taking them offline

Outcome: Take founder strategy debates offline so the team doesn't misread healthy conflict as instability.

Context: Narang and co-founder Steve debated strategy openly while sitting on the floor among 25-40 people, who asked "why are you guys fighting?" They learned the hard way to take disagreements away from the team.

if we've got disagreements to work through, we're gonna go away so people are distracted by what We're doing.
Aman Narang
Ongoing, from the first ~25-person stage per
  1. 1

  2. 2

  3. 3

  4. 4

  5. 5

  6. 6

Before you start

  • · High-trust co-founder relationship that can debate hard
  • · Awareness of how the team reads founder body language
org-design

Pivot the sales conversation from your feature to their whole stack to find the real pain

Outcome: When a feature pitch gets no time, reframe discovery to the customer's whole stack to find the platform-sized pain.

Context: Toast got barely any time pitching phone checkout, but when they asked restaurateurs broadly about their technology, the legacy-platform frustration ("oil and water") poured out — revealing the full-stack opportunity.

when we pivoted the conversation to talk about like, Hey, how do you like your technology more broadly? Broadly. And a couple teams in merge one didn't, none of them liked their legacy platforms.
Aman Narang
Weeks to a few months of discovery before re-scoping per
  1. 1

  2. 2

  3. 3

  4. 4

  5. 5

  6. 6

  7. 7

Before you start

  • · Direct founder access to target customers
  • · Willingness to abandon the original feature framing
go-to-market

Land-and-expand the platform via acquisitions and organic builds pulled by customers

Outcome: Expand the platform via a mix of organic builds and acquisitions, sequenced by customer pull.

Context: From the POS+payments base, Toast added capital, online ordering, gift/loyalty, CRM organically, and acquired companies for payroll, scheduling, back-office inventory, and AP automation. The sequence was driven by what customers were pulling for (e.g. capital, because they wouldn't switch off legacy without it).

We bought a company for payroll and scheduling separate companies for payroll and scheduling. We bought a company for back office inventory and AP automation. And, and of course we continue to build organically in the platform
Aman Narang
Growth-to-scale phase per
  1. 1

  2. 2

  3. 3

  4. 4

  5. 5

  6. 6

  7. 7

Before you start

  • · Installed, trusted core platform
  • · Capital for acquisitions
  • · Integration capability to unify reporting
product

Ship a card-on-file walk-out flow by connecting reservations to the POS guest profile

Outcome: Capture card-on-file at booking and connect it to the POS to let guests walk out, hotel-checkout style.

Context: Toast launched a walk-out flow: book/waitlist through Toast Tables (integrating Resi, OpenTable, etc.) with a card on file, then leave without the pay-and-wait friction. The same guest-profile-in-POS connection enables pre-firing the kitchen on check-in and personalized recommendations.

imagine, right? You get a table where you join a wait list, you've got a card on file. So toast now just like, just launched that. And so when you go to Toast Restaurant and you book through Toast Tables soon with resi and talk and others, you're gonna be able to walk out.
Aman Narang
Newly launched; expanding via reservation partners per
  1. 1

  2. 2

  3. 3

  4. 4

  5. 5

  6. 6

  7. 7

Before you start

  • · Reservation/waitlist integrations
  • · Card-on-file capture at booking
  • · Guest profile connected to POS
product

Stretch a small seed by releasing it in tranches and weighting comp to equity

Outcome: Release seed capital in milestone tranches and weight comp to equity to extend a tiny runway.

Context: Steve Papa committed $500K but released it $100K at a time (and required approving every hire). Toast took no salaries and leaned comp toward equity to make the $500K count, not knowing if more capital would come.

He said, I'm gonna commit 500 K and the a hundred K showed up in our account. And I said, I thought you said 500 K. And, and he said, well, I'm gonna view 500 KA hundred K at a time
Aman Narang
Pre-institutional round (here ~2012-2015) per
  1. 1

  2. 2

  3. 3

  4. 4

  5. 5

  6. 6

  7. 7

Before you start

  • · An investor willing to tranche and engage hands-on
  • · Team that accepts equity-heavy comp
  • · Genuine frugality discipline
fundraising

Codify mission, purpose and values with a cross-section of employees, then make them true

Outcome: Have top employees author mission/values, then wire them into hiring and onboarding to make them true.

Context: When the incoming CEO joined (2015), Toast took top employees and had them define mission, purpose, and values, plus basic hire/onboard infrastructure. Narang stresses values must be made true, not just put on a wall — and credits this as a bigger driver than strategy.

we asked him like, what makes toast tick? And we had them put together our mission and our purpose and our values and our, you know, we, we put some basic infrastructure in place for how we hire an onboard. And that made a huge difference.
Aman Narang
Triggered at the scaling inflection (~50-100+ people) per
  1. 1

  2. 2

  3. 3

  4. 4

  5. 5

  6. 6

  7. 7

Before you start

  • · A leader who can convene the team without making it their show
  • · Commitment to operationalize, not just publish, the values
org-design

Deliberately serve every sub-segment to force a universal platform and a fundable sales motion

Outcome: Intentionally serve every sub-segment to force a universal platform and a sales team with enough TAM to be fundable.

Context: Toast's first 10-20 customers spanned food trucks, cafes, quick-serve, full-serve, fine dining, a bar, and a nightclub (with pool-table billing). The deliberate breadth built universal capability and a sales motion that could address the whole industry — not a narrow slice.

we're gonna build this platform that can work universally across these different restaurant types... if you're gonna hire a sales team, they gotta be able to sell, not just to a small segment of the restaurant industry, gotta sell across different restaurant types.
Aman Narang
Initial PMF phase (months 9-24) per
  1. 1

  2. 2

  3. 3

  4. 4

  5. 5

  6. 6

  7. 7

Before you start

  • · Conviction to resist 'why don't you focus' pressure
  • · Capital/runway to build broad capability
go-to-market

Decision Moments

Actual decisions, real outcomes

Specific decisions narrated in the episode with their outcomes and transferable lessons.

Toast's phone-checkout app had nine months of no traction because restaurants ran on disconnected 1990s on-prem systems with no APIs to integrate with — there was no path to scale the original product.

Did: Instead of pushing the checkout feature, they pivoted the customer conversation to "how do you like your technology more broadly," discovered universal frustration with fragmented legacy stacks, and re-scoped from a checkout app to a full restaurant operating system — deliberately serving every restaurant sub-segment and leaning into complexity.Outcome: Found the beginnings of product-market fit; the full platform launched summer 2013 and reached ~$2M ARR by end of 2014. The complexity-as-moat thesis became the company's enduring strategy and later carried into retail.

A failed first product is a probe. The durable business is often the larger pain the failure reveals in customer conversation — and leaning into the vertical's complexity, rather than narrowing focus, can be the moat.

Part of an emerging decision pattern across multiple episodes

By 2014 Toast had strong customer pull but a wide performance distribution, worsening financials as it added salespeople, weak customer sentiment, and no org foundations (hiring, onboarding, assessment, culture). It was the closest the company came to not working. First-time founders had never managed at scale.

Did: Brought in Chris (who had scaled customer success at Endeca) as CEO in 2015; the founders stepped back to president roles. Had top employees codify mission, purpose, and values and built basic hire/onboard infrastructure.Outcome: Chris ran the company 2015-2023 (~8 years); the move worked. Toast built the org foundations, scaled to a ~$20B public company and 6,500 employees, and Narang took the CEO seat back end of 2023.

When first-time founders hit the scaling wall, importing a proven operator to install org systems — even at the cost of the CEO title — can be the unlock. The long outside-CEO apprenticeship is a feature, not a failure.

Part of an emerging decision pattern across multiple episodes

Around 2021, post-IPO, Toast felt it had lost some entrepreneurial spirit, and faced the IPO-era question of how big the real TAM was beyond SMB US restaurants (where it had ~5% share at IPO, ~20% now).

Did: Launched New Ventures — a structurally separate internal incubator (isolated teams, focused comp plans, entrepreneur hires) echoing the Endeca "special operations" team Toast was born from — to find new product-market fit, and sized local commerce ($7-8T) as a 10x+ TAM beyond restaurants.Outcome: New Ventures shipped Toast Retail (potentially bigger than restaurants long-term), guest-facing reservations/waitlist/Toast Local, catering/events, and advertising — with a high hit rate, restoring zero-to-one capability and opening a 10x TAM.

Scale operators and zero-to-one builders are different people doing different work; a scaled company needs a protected, structurally separate unit to keep finding new product-market fit and new growth vectors.

Part of an emerging decision pattern across multiple episodes

Tensions surfaced

Contradictions and trade-offs the episode raises — judgment calls a thoughtful operator has to navigate.

Tension

Slow down to fix foundations vs keep the revenue momentum that keeps you alive

Momentum is sacred, but pause it deliberately when scaling would burn the foundations the momentum depends on.

Toast felt intense pressure to "put the pedal down" going from 10 to 50 customers (a literal thermometer chart for morale), yet the software didn't work offline, hardware was failing, and support was a Google Voice line. Co-founders and investor forced a deliberate slowdown to fix foundations, then re-accelerated — resolving the tension via a temporary readiness gate.

When momentum and readiness conflict, gate growth temporarily — pause precisely when scaling would destroy the foundations momentum needs.

Corpus connection

Where this episode fits for retrieval

What kinds of decisions this briefing is best pulled into.

Primary decisions

  • strategic-bet
  • hire
  • market-entry
  • pricing