Marble growth playbook

Publish five pages at /solutions/<segment>, starting with Franchise Groups and Multi-unit/Hospitality Groups, where Marble sells first. Each page covers the segment's back-of-house problem in its own words (franchise: repeat counts across stores and consistent pars; cafes: milk, pastry and perishables waste; bars: liquor variance and pour costs), the modules that fix it, the matching customer logos and A&W numbers, the integrations that segment uses (for example 7shifts or HotSchedules for labor, OpenTable, Resy or SevenRooms for hospitality), and a demo form that records the segment. Point the footer links at the pages, add them to the sitemap, and give each a title and description built around the segment search term.

Segment pages turn a generic pitch into "built for operators like me", which matters most to multi-unit and franchise buyers comparing Marble with long-established rivals. They also give search engines and AI assistants pages to rank and quote for segment searches, which a 9-URL site can't do, and they tell the demo call which story to tell before it starts.

[ ] /solutions/franchise-groups with the A&W 6-day timeline
[ ] /solutions/hospitality-groups with reservations and labor integrations
[ ] /solutions/independent-restaurants, /cafes, /bars
[ ] Footer and nav links point to the pages, not /get-started
[ ] Pages added to sitemap.xml
[ ] Unique title and meta description per segment
[ ] Matching customer logos on each page
[ ] Demo form records the segment

• Thin proof for some segments

  Launch Franchise and Hospitality first, and add a story to the others as idea 2 delivers them.

• Pages read as generic

  Use one concrete workflow per segment, in operators' own words.

• Your websiteSite structure (checked November 10, 2026): the sitemap lists 9 URLs (home, get-started, careers, about, case-studies, one case study, impact-calculator, privacy, terms). The five footer Solutions links (Independent Restaurants, Cafes, Bars, Franchise Groups, Hospitality Groups) all point to /get-started; the Resources menu has no items; Integrations is a homepage anchor; there is no blog, pricing, security or comparison page; /refer exists but is not in the sitemap.Source
• Your websiteCase studies: one published story (A&W, July 5, 2026, a multi-unit QSR franchise) next to a "Coming soon" placeholder. Results: 97% item-level forecast accuracy, 90% faster inventory counts, 6 days to full onboarding for the first site and later sites live in hours; the quote is attributed to a "Multi-Unit Franchise Owner"; no location count or dollar savings.Source
• Your websiteHomepage: headline "The autonomous back of house for restaurants." Four modules (Inventory, Labor, Finance, Intelligence): voice and vision counts, an AI assistant, autopilot agents with human approval, demand forecasting, agentic ordering, prep and scheduling. Eight customer logos including A&W Restaurants, The Kati Roll Company, Corgi Cafe, Ess-a-Bagel and Kitchen & Coffee. Integrations shown as logos only: Toast, Square, Clover, 7shifts, Sling, HotSchedules, ezCater, OpenTable, Resy, SevenRooms, QuickBooks, Xero, Restaurant365, Gusto, ADP. No pricing; every main call to action is "Get a demo".Source
• Y Combinator profileSummer 2026 batch, founded 2026, team of 3, New York. Launch claims: 90% faster counts, 97% forecast accuracy, a 40% reduction in food waste and 30+ labor hours saved per week for managers. Marble targets multi-unit restaurant groups first.Source

Write four short stories in the A&W format, one each for a cafe, a bar or hospitality group, an independent restaurant and a second multi-unit group, picked from customers already on the logo wall where possible. Each story gives location count, go-live days, count time before and after, and one money figure (waste or labor overage), with a named role and a short phone video. Make story collection part of the 30-day deployment checklist the Founding Ops hire owns.

Early-stage buyers buy on peer proof, and one story can't cover five segments. Building it into deployment means each new restaurant becomes a story without a separate project.

• Customers decline to be named

  Offer an anonymised story by segment and location count.

• Your websiteCase studies: one published story (A&W, July 5, 2026, a multi-unit QSR franchise) next to a "Coming soon" placeholder. Results: 97% item-level forecast accuracy, 90% faster inventory counts, 6 days to full onboarding for the first site and later sites live in hours; the quote is attributed to a "Multi-Unit Franchise Owner"; no location count or dollar savings.Source
• HiringFounding Ops (New York, in office, $100K-$150K, 0.20%-0.75% equity): owns deployments from contract signing to go-live, is the primary customer contact, builds repeatable playbooks and supports sales through calls, pilots and multi-unit rollouts. The post says Marble signed 150+ restaurants in under four months across the US, Canada and Europe, growing 25% week over week for three months, with a team of 3.Source
• Competitor researchNory calls itself "the agentic AI restaurant operating system", also claims 97% demand forecast accuracy, has an Ordering Assistant in beta and an ROI calculator, and serves independent, franchise, multi-location and enterprise groups. Crunchtime cites 850+ restaurant brands and 150,000+ locations. MarketMan publishes pricing ($249, $299, from $449 a month) for 15,000+ restaurants.See the research Source

Render the final figures in the HTML and animate from there, or drop the animation. Add a short static summary to the impact calculator page (the inputs, the assumptions and an example result for a 5-location group) so it can be read, shared and indexed, and let a visitor email the result to a partner.

The ROI numbers are Marble's strongest sales argument. Making them readable everywhere costs a few lines of code and stops crawlers and AI summaries from recording zeros.

• The figures look less dramatic without animation

  Keep a light count-up that starts from the true value's neighbourhood.

• Your websiteROI and demo pages: the homepage ROI section's six figures (manual hours saved per location per week, value of managed inventory, food waste reduction, labor overage reduction, count time saved) read 0 in the page text on two separate reads; the impact calculator page text is only "See what impact Marble could have on your business" and "Click anywhere to begin"; the demo page says "Book your demo for free" and "Pick a time".Source
• Y Combinator profileSummer 2026 batch, founded 2026, team of 3, New York. Launch claims: 90% faster counts, 97% forecast accuracy, a 40% reduction in food waste and 30+ labor hours saved per week for managers. Marble targets multi-unit restaurant groups first.Source

Serve the hero banner as a properly sized AVIF/WebP with a preload hint and responsive sizes, lazy-load the logo carousel and lower images, and remove unused JavaScript and render-blocking font requests. Re-test until the homepage passes in real-user data.

Restaurant operators often browse on phones between shifts, so a faster first screen keeps them on the page. Passing Core Web Vitals also helps search ranking for the new segment pages.

• Hero looks softer

  Compare at high-density screens before shipping.

• PageSpeedMobile test (November 10, 2026): performance 67 (desktop 72); the homepage downloads 10.3 MB, 6.1 MB of it images, with about 5 MB of image savings available; the largest element is the hero banner image, painting at 21.4 s in the lab. Real Chrome visitors to the homepage fail Core Web Vitals on LCP (p75 2.8 s against a 2.5 s target) while INP (189 ms) and CLS (0) pass; across the whole site LCP is 2.0 s and 12% faster than before.Source

Add a customer version of the program: a referral link in the product and in the 30-day check-in, a reward that fits operators (a month free per location, or the cash), and a "refer a fellow franchisee" message for franchise customers. Add /refer to the sitemap and link it from the footer of every deployment email.

Happy customers in a franchise system bring warm, look-alike leads who already trust the peer, and the 30-day threshold keeps rewards tied to real deployments.

• Franchise rules on vendor promotion

  Let franchisees share privately rather than through the franchisor.

• Your websiteReferral page: $500 on a prepaid card or an Oura Ring 5 ("About $569 in value") per referred restaurant, earned after 30 days of active deployment; rewards ship in the US; referrers enter a name and email to get a link; the homepage pitches it to non-operators.Source
• Your websiteSite structure (checked November 10, 2026): the sitemap lists 9 URLs (home, get-started, careers, about, case-studies, one case study, impact-calculator, privacy, terms). The five footer Solutions links (Independent Restaurants, Cafes, Bars, Franchise Groups, Hospitality Groups) all point to /get-started; the Resources menu has no items; Integrations is a homepage anchor; there is no blog, pricing, security or comparison page; /refer exists but is not in the sitemap.Source
• Your websiteCase studies: one published story (A&W, July 5, 2026, a multi-unit QSR franchise) next to a "Coming soon" placeholder. Results: 97% item-level forecast accuracy, 90% faster inventory counts, 6 days to full onboarding for the first site and later sites live in hours; the quote is attributed to a "Multi-Unit Franchise Owner"; no location count or dollar savings.Source

Keep the usetruffle.ai redirect permanently, with page-to-page redirects for any old paths. Update every directory and profile (YC, startup databases, launch posts, LinkedIn) to Marble with a "formerly Truffle" line, and add "Marble (formerly Truffle)" to the About page so the two names are connected.

It moves the old name's links and mentions to the new domain and stops buyers or investors from finding two half-companies.

• Letting the old domain lapse

  Renew usetruffle.ai for several years.

• Your websiteRebrand: the company was called Truffle; usetruffle.ai returns a 308 permanent redirect to joinmarble.ai; joinmarble.ai was registered on July 23, 2026; a startup directory page is still titled truffle-yc-s26.Source

Add a short "How we measure" note for each claim: which customers, how many locations, what period, and the metric definition (for example, the forecast error measure behind 97%). Use one consistent set of numbers on the homepage, case studies and YC profile, and lead with the claims rivals can't match, such as count time and go-live days.

Group finance buyers discount unsourced claims; a method note makes the same number believable and turns a shared 97% into a fair comparison Marble can win.

• Numbers fall when sourced

  Publish the honest figure; credibility is worth more than a point.

• Y Combinator profileSummer 2026 batch, founded 2026, team of 3, New York. Launch claims: 90% faster counts, 97% forecast accuracy, a 40% reduction in food waste and 30+ labor hours saved per week for managers. Marble targets multi-unit restaurant groups first.Source
• Competitor researchNory calls itself "the agentic AI restaurant operating system", also claims 97% demand forecast accuracy, has an Ordering Assistant in beta and an ROI calculator, and serves independent, franchise, multi-location and enterprise groups. Crunchtime cites 850+ restaurant brands and 150,000+ locations. MarketMan publishes pricing ($249, $299, from $449 a month) for 15,000+ restaurants.See the research Source
• PressLaunch coverage: Runtime Wire (July 30, 2026) named no customers and grouped Marble with MarketMan, Stockcount, Crunchtime and MarginEdge; a restaurant-tech blog (July 10, 2026) repeated Marble's figures and said new locations typically go live within two weeks. No independent reviews were found.Source

Publish a plain-text facts page (and an llms.txt pointing to it): what Marble does, modules, integrations, segments, onboarding time, sourced results, how agents get approval, and "formerly Truffle". Link it from the footer next to the Ask AI links, and keep it current.

Buyers increasingly ask assistants for tool shortlists, and a clear, sourced page is what those answers draw on. It also makes the existing Ask AI links return accurate answers.

• Assistants update slowly

  Also fix the directory listings in idea 10.

• Your websiteSite structure (checked November 10, 2026): the sitemap lists 9 URLs (home, get-started, careers, about, case-studies, one case study, impact-calculator, privacy, terms). The five footer Solutions links (Independent Restaurants, Cafes, Bars, Franchise Groups, Hospitality Groups) all point to /get-started; the Resources menu has no items; Integrations is a homepage anchor; there is no blog, pricing, security or comparison page; /refer exists but is not in the sitemap.Source
• Your websiteROI and demo pages: the homepage ROI section's six figures (manual hours saved per location per week, value of managed inventory, food waste reduction, labor overage reduction, count time saved) read 0 in the page text on two separate reads; the impact calculator page text is only "See what impact Marble could have on your business" and "Click anywhere to begin"; the demo page says "Book your demo for free" and "Pick a time".Source
• Your websiteRebrand: the company was called Truffle; usetruffle.ai returns a 308 permanent redirect to joinmarble.ai; joinmarble.ai was registered on July 23, 2026; a startup directory page is still titled truffle-yc-s26.Source

Build an integration page for each of the top POS systems and labor and accounting tools: what data flows, what Marble does with it, setup time, and a demo form tagged with the POS. Then apply to each partner's app marketplace with the same copy.

Integration pages catch high-intent searches like "Toast inventory app" and are what marketplace listings link to. They also answer the first question every operator asks on a demo.

• Overstating integration depth

  List exactly which data syncs, and how often.

• #1 Build real pages behind the five Solutions links, which today all open the demo form

• Your websiteHomepage: headline "The autonomous back of house for restaurants." Four modules (Inventory, Labor, Finance, Intelligence): voice and vision counts, an AI assistant, autopilot agents with human approval, demand forecasting, agentic ordering, prep and scheduling. Eight customer logos including A&W Restaurants, The Kati Roll Company, Corgi Cafe, Ess-a-Bagel and Kitchen & Coffee. Integrations shown as logos only: Toast, Square, Clover, 7shifts, Sling, HotSchedules, ezCater, OpenTable, Resy, SevenRooms, QuickBooks, Xero, Restaurant365, Gusto, ADP. No pricing; every main call to action is "Get a demo".Source
• Your websiteSite structure (checked November 10, 2026): the sitemap lists 9 URLs (home, get-started, careers, about, case-studies, one case study, impact-calculator, privacy, terms). The five footer Solutions links (Independent Restaurants, Cafes, Bars, Franchise Groups, Hospitality Groups) all point to /get-started; the Resources menu has no items; Integrations is a homepage anchor; there is no blog, pricing, security or comparison page; /refer exists but is not in the sitemap.Source
• HiringFounding Engineer (New York, in office, $150K-$225K, 1%-3% equity): owns the agent harness, including human approval for purchasing decisions; builds POS, accounting and payroll integrations; makes agents reliable through evals and failure recovery.Source

Publish one fair page per rival: what each does well, where Marble differs (voice and photo counts, agents that act with approval, days-not-months go-live), price model, setup time and a switching plan that imports existing recipes, pars and vendors. Date each page and cite the rival's own site.

Comparison searches come from buyers already in the market. A new brand gets onto the shortlist by being on the pages buyers read while comparing.

• Out-of-date claims about rivals

  Date the page and link to sources.

• PressLaunch coverage: Runtime Wire (July 30, 2026) named no customers and grouped Marble with MarketMan, Stockcount, Crunchtime and MarginEdge; a restaurant-tech blog (July 10, 2026) repeated Marble's figures and said new locations typically go live within two weeks. No independent reviews were found.Source
• Competitor researchNory calls itself "the agentic AI restaurant operating system", also claims 97% demand forecast accuracy, has an Ordering Assistant in beta and an ROI calculator, and serves independent, franchise, multi-location and enterprise groups. Crunchtime cites 850+ restaurant brands and 150,000+ locations. MarketMan publishes pricing ($249, $299, from $449 a month) for 15,000+ restaurants.See the research Source
• Competitor researchMarginEdge: "AI runs the numbers. You run the restaurant.", a published $350 per month per location, "Trusted by 12,000+ restaurants", distributor integrations such as US Foods, Gordon Food Service and Shamrock, setup within a few weeks. Restaurant365: 52,000+ restaurants, R365 AI with AI Scheduling in early access, and a mobile app in English and Spanish.See the research Source
• Your websiteSite structure (checked November 10, 2026): the sitemap lists 9 URLs (home, get-started, careers, about, case-studies, one case study, impact-calculator, privacy, terms). The five footer Solutions links (Independent Restaurants, Cafes, Bars, Franchise Groups, Hospitality Groups) all point to /get-started; the Resources menu has no items; Integrations is a homepage anchor; there is no blog, pricing, security or comparison page; /refer exists but is not in the sitemap.Source

Add a short pricing page with a "from $X per location per month" line, what is included (modules, onboarding, agents), volume terms for groups, and the go-live promise. Keep a custom quote for groups above a set number of locations.

A starting price qualifies demos, removes a reason to leave for a rival's page and gives the comparison pages something concrete to show.

• Locks in a low price

  Publish a "from" price and keep group quotes custom.

• Competitor researchMarginEdge: "AI runs the numbers. You run the restaurant.", a published $350 per month per location, "Trusted by 12,000+ restaurants", distributor integrations such as US Foods, Gordon Food Service and Shamrock, setup within a few weeks. Restaurant365: 52,000+ restaurants, R365 AI with AI Scheduling in early access, and a mobile app in English and Spanish.See the research Source
• Competitor researchNory calls itself "the agentic AI restaurant operating system", also claims 97% demand forecast accuracy, has an Ordering Assistant in beta and an ROI calculator, and serves independent, franchise, multi-location and enterprise groups. Crunchtime cites 850+ restaurant brands and 150,000+ locations. MarketMan publishes pricing ($249, $299, from $449 a month) for 15,000+ restaurants.See the research Source
• Your websiteHomepage: headline "The autonomous back of house for restaurants." Four modules (Inventory, Labor, Finance, Intelligence): voice and vision counts, an AI assistant, autopilot agents with human approval, demand forecasting, agentic ordering, prep and scheduling. Eight customer logos including A&W Restaurants, The Kati Roll Company, Corgi Cafe, Ess-a-Bagel and Kitchen & Coffee. Integrations shown as logos only: Toast, Square, Clover, 7shifts, Sling, HotSchedules, ezCater, OpenTable, Resy, SevenRooms, QuickBooks, Xero, Restaurant365, Gusto, ADP. No pricing; every main call to action is "Get a demo".Source

For each franchise brand with a live franchisee, build a brand kit: pre-mapped menu, recipes, pars and approved vendors, so a new franchisee in the same system goes live in a day. Offer it at franchisee meetings and in brand-specific outreach, with the live franchisee's results (with permission).

Reused setup cuts the cost and time of every extra deal, and peers in the same system trust each other's numbers.

• Brand data belongs to the franchisor

  Use only each franchisee's own data and get consent before reuse.

• #2 Fill the empty case-study slot: one published story per segment from the 150+ signed restaurants

• Your websiteCase studies: one published story (A&W, July 5, 2026, a multi-unit QSR franchise) next to a "Coming soon" placeholder. Results: 97% item-level forecast accuracy, 90% faster inventory counts, 6 days to full onboarding for the first site and later sites live in hours; the quote is attributed to a "Multi-Unit Franchise Owner"; no location count or dollar savings.Source
• HiringFounding Ops (New York, in office, $100K-$150K, 0.20%-0.75% equity): owns deployments from contract signing to go-live, is the primary customer contact, builds repeatable playbooks and supports sales through calls, pilots and multi-unit rollouts. The post says Marble signed 150+ restaurants in under four months across the US, Canada and Europe, growing 25% week over week for three months, with a team of 3.Source
• Y Combinator profileSummer 2026 batch, founded 2026, team of 3, New York. Launch claims: 90% faster counts, 97% forecast accuracy, a 40% reduction in food waste and 30+ labor hours saved per week for managers. Marble targets multi-unit restaurant groups first.Source

Post one short piece a week: a real count filmed in a kitchen, a waste or labor lesson from Marble's data, a vendor price swing. Publish it on a company LinkedIn page and in a monthly email to prospects and customers, and keep the best ones as resources on the site.

Founder content builds trust fast at this stage, and a company page and email list keep the audience when the team grows.

• Founder time is scarce

  Batch-record a month of clips in one afternoon.

• Your websiteAbout page: Marble is building the agentic operating system for physical-world industries, "starting with restaurants", organised as perceive, reason and act, with an agent named Remy. The CEO grew up running a family restaurant business and covered the restaurant sector as an investment banker; the CTO led AI and machine learning for electronic trading. The site's social links go to the CEO's personal LinkedIn and X; no company blog, newsletter or YouTube channel was found.Source
• Your websiteSite structure (checked November 10, 2026): the sitemap lists 9 URLs (home, get-started, careers, about, case-studies, one case study, impact-calculator, privacy, terms). The five footer Solutions links (Independent Restaurants, Cafes, Bars, Franchise Groups, Hospitality Groups) all point to /get-started; the Resources menu has no items; Integrations is a homepage anchor; there is no blog, pricing, security or comparison page; /refer exists but is not in the sitemap.Source

Publish the day-by-day plan (day 1 POS, days 2-4 menus, recipes, vendors and pars, days 5-6 agents) and guarantee it for single and multi-unit customers who meet a short checklist (POS access, a vendor list, one manager hour a day). Show the guarantee on the demo page and in proposals.

A specific, backed promise turns Marble's speed into a reason to sign now, and the checklist protects the team from the cases it can't control.

• Messy POS data slows go-live

  Start the clock only when the checklist is complete.

• Your websiteCase studies: one published story (A&W, July 5, 2026, a multi-unit QSR franchise) next to a "Coming soon" placeholder. Results: 97% item-level forecast accuracy, 90% faster inventory counts, 6 days to full onboarding for the first site and later sites live in hours; the quote is attributed to a "Multi-Unit Franchise Owner"; no location count or dollar savings.Source
• HiringFounding Ops (New York, in office, $100K-$150K, 0.20%-0.75% equity): owns deployments from contract signing to go-live, is the primary customer contact, builds repeatable playbooks and supports sales through calls, pilots and multi-unit rollouts. The post says Marble signed 150+ restaurants in under four months across the US, Canada and Europe, growing 25% week over week for three months, with a team of 3.Source
• Competitor researchMarginEdge: "AI runs the numbers. You run the restaurant.", a published $350 per month per location, "Trusted by 12,000+ restaurants", distributor integrations such as US Foods, Gordon Food Service and Shamrock, setup within a few weeks. Restaurant365: 52,000+ restaurants, R365 AI with AI Scheduling in early access, and a mobile app in English and Spanish.See the research Source

Publish /trust: which actions need approval, spending limits per location, audit logs, who can see what across locations, where data is stored, how POS and payroll credentials are handled, and how customer data is (or isn't) used for model training. Add a one-page PDF for finance and IT reviewers.

Group buyers have a finance or IT reviewer, and a clear page shortens that step instead of starting a long questionnaire.

• Promising more than is built

  Only list controls that exist today, and mark planned ones.

• HiringFounding Engineer (New York, in office, $150K-$225K, 1%-3% equity): owns the agent harness, including human approval for purchasing decisions; builds POS, accounting and payroll integrations; makes agents reliable through evals and failure recovery.Source
• Your websiteSite structure (checked November 10, 2026): the sitemap lists 9 URLs (home, get-started, careers, about, case-studies, one case study, impact-calculator, privacy, terms). The five footer Solutions links (Independent Restaurants, Cafes, Bars, Franchise Groups, Hospitality Groups) all point to /get-started; the Resources menu has no items; Integrations is a homepage anchor; there is no blog, pricing, security or comparison page; /refer exists but is not in the sitemap.Source
• Competitor researchMarginEdge: "AI runs the numbers. You run the restaurant.", a published $350 per month per location, "Trusted by 12,000+ restaurants", distributor integrations such as US Foods, Gordon Food Service and Shamrock, setup within a few weeks. Restaurant365: 52,000+ restaurants, R365 AI with AI Scheduling in early access, and a mobile app in English and Spanish.See the research Source

Offer a one-week trial on one location: connect the POS, set up counts for the top 50 items, and run daily voice or photo counts with a time-per-count report at the end. Convert to a full rollout with the week's data already loaded.

Kitchen staff using it for a week is stronger proof than any demo, and the setup work carries into the paid rollout.

• Too many low-intent trials

  Offer it after a demo, to groups with 3+ locations.

• Your websiteCase studies: one published story (A&W, July 5, 2026, a multi-unit QSR franchise) next to a "Coming soon" placeholder. Results: 97% item-level forecast accuracy, 90% faster inventory counts, 6 days to full onboarding for the first site and later sites live in hours; the quote is attributed to a "Multi-Unit Franchise Owner"; no location count or dollar savings.Source
• Your websiteROI and demo pages: the homepage ROI section's six figures (manual hours saved per location per week, value of managed inventory, food waste reduction, labor overage reduction, count time saved) read 0 in the page text on two separate reads; the impact calculator page text is only "See what impact Marble could have on your business" and "Click anywhere to begin"; the demo page says "Book your demo for free" and "Pick a time".Source

Offer accountants a partner tier: a multi-client view, clean invoice data in their ledger, a revenue share or per-restaurant fee, and co-branded onboarding. Recruit through restaurant-focused accounting firms and their newsletters.

One accountant can bring many restaurants, and clean data saves the accountant hours each month, so the partner benefits even without the fee.

• Accountants favour their current tools

  Lead with the export to their existing ledger.

• #8 Open referrals to operators and franchisees, not just people outside the industry

• Your websiteHomepage: headline "The autonomous back of house for restaurants." Four modules (Inventory, Labor, Finance, Intelligence): voice and vision counts, an AI assistant, autopilot agents with human approval, demand forecasting, agentic ordering, prep and scheduling. Eight customer logos including A&W Restaurants, The Kati Roll Company, Corgi Cafe, Ess-a-Bagel and Kitchen & Coffee. Integrations shown as logos only: Toast, Square, Clover, 7shifts, Sling, HotSchedules, ezCater, OpenTable, Resy, SevenRooms, QuickBooks, Xero, Restaurant365, Gusto, ADP. No pricing; every main call to action is "Get a demo".Source
• Your websiteReferral page: $500 on a prepaid card or an Oura Ring 5 ("About $569 in value") per referred restaurant, earned after 30 days of active deployment; rewards ship in the US; referrers enter a name and email to get a link; the homepage pitches it to non-operators.Source

Send each owner a weekly one-screen summary per location: count time, waste flagged, orders approved, labor against forecast and an estimated dollar figure, plus one suggestion. Flag locations where counts stopped, so Ops can step in before usage drops.

It keeps the buyer seeing the return, catches stalled locations early and makes expansion and referral asks easy.

• Estimated savings look inflated

  Show the method and keep estimates conservative.

• Your websiteROI and demo pages: the homepage ROI section's six figures (manual hours saved per location per week, value of managed inventory, food waste reduction, labor overage reduction, count time saved) read 0 in the page text on two separate reads; the impact calculator page text is only "See what impact Marble could have on your business" and "Click anywhere to begin"; the demo page says "Book your demo for free" and "Pick a time".Source
• Your websiteReferral page: $500 on a prepaid card or an Oura Ring 5 ("About $569 in value") per referred restaurant, earned after 30 days of active deployment; rewards ship in the US; referrers enter a name and email to get a link; the homepage pitches it to non-operators.Source
• Y Combinator profileSummer 2026 batch, founded 2026, team of 3, New York. Launch claims: 90% faster counts, 97% forecast accuracy, a 40% reduction in food waste and 30+ labor hours saved per week for managers. Marble targets multi-unit restaurant groups first.Source

At day 30 of the first location, present that location's results and a fixed offer for the rest of the group: next-day go-live per location and a volume price. Make it a standard step in the Founding Ops deployment playbook.

Results from their own store are the strongest argument for the rest of the group, and it lifts revenue per customer without new acquisition cost.

• Discounts erode price

  Tie the discount to a group-wide commitment.

• #19 A Monday email to every owner: what Marble saved each location last week

• Your websiteCase studies: one published story (A&W, July 5, 2026, a multi-unit QSR franchise) next to a "Coming soon" placeholder. Results: 97% item-level forecast accuracy, 90% faster inventory counts, 6 days to full onboarding for the first site and later sites live in hours; the quote is attributed to a "Multi-Unit Franchise Owner"; no location count or dollar savings.Source
• HiringFounding Ops (New York, in office, $100K-$150K, 0.20%-0.75% equity): owns deployments from contract signing to go-live, is the primary customer contact, builds repeatable playbooks and supports sales through calls, pilots and multi-unit rollouts. The post says Marble signed 150+ restaurants in under four months across the US, Canada and Europe, growing 25% week over week for three months, with a team of 3.Source

Offer a free tool: an operator forwards or uploads a month of invoices and gets a report of price changes by item and vendor, plus weekly alerts. The report ends with what Marble's ordering agent would have done. Prospects who hit a set savings figure get a call.

It lets operators see value from their own data before a demo, and every upload shows what they buy and from whom, so the sales call starts from real numbers.

• Free users never upgrade

  Cap the free report to one month and one location.

• Your websiteHomepage: headline "The autonomous back of house for restaurants." Four modules (Inventory, Labor, Finance, Intelligence): voice and vision counts, an AI assistant, autopilot agents with human approval, demand forecasting, agentic ordering, prep and scheduling. Eight customer logos including A&W Restaurants, The Kati Roll Company, Corgi Cafe, Ess-a-Bagel and Kitchen & Coffee. Integrations shown as logos only: Toast, Square, Clover, 7shifts, Sling, HotSchedules, ezCater, OpenTable, Resy, SevenRooms, QuickBooks, Xero, Restaurant365, Gusto, ADP. No pricing; every main call to action is "Get a demo".Source
• Your websiteROI and demo pages: the homepage ROI section's six figures (manual hours saved per location per week, value of managed inventory, food waste reduction, labor overage reduction, count time saved) read 0 in the page text on two separate reads; the impact calculator page text is only "See what impact Marble could have on your business" and "Click anywhere to begin"; the demo page says "Book your demo for free" and "Pick a time".Source

Publish an anonymised quarterly report: median food waste, count frequency, labor overage and price inflation by segment and location count, with a free "how do you compare" form. Pitch it to restaurant trade press.

Original data earns press, links and AI-assistant citations, and the comparison form produces qualified leads.

• Sample too small or skewed

  State sample sizes and only publish segments with enough locations.

• #21 Back every headline number with a source and a method note, starting with the 97% that Nory also claims

• HiringFounding Ops (New York, in office, $100K-$150K, 0.20%-0.75% equity): owns deployments from contract signing to go-live, is the primary customer contact, builds repeatable playbooks and supports sales through calls, pilots and multi-unit rollouts. The post says Marble signed 150+ restaurants in under four months across the US, Canada and Europe, growing 25% week over week for three months, with a team of 3.Source
• PressLaunch coverage: Runtime Wire (July 30, 2026) named no customers and grouped Marble with MarketMan, Stockcount, Crunchtime and MarginEdge; a restaurant-tech blog (July 10, 2026) repeated Marble's figures and said new locations typically go live within two weeks. No independent reviews were found.Source
• Competitor researchNory calls itself "the agentic AI restaurant operating system", also claims 97% demand forecast accuracy, has an Ordering Assistant in beta and an ROI calculator, and serves independent, franchise, multi-location and enterprise groups. Crunchtime cites 850+ restaurant brands and 150,000+ locations. MarketMan publishes pricing ($249, $299, from $449 a month) for 15,000+ restaurants.See the research Source

Start with the one or two distributors most of Marble's customers use: live catalogue prices, order submission after approval and invoice matching. Then pitch distributors on co-marketing to their own restaurant accounts, since accurate forecast-based orders mean fewer returns and emergency deliveries.

Closing the loop from forecast to delivered order is the strongest version of the "autonomous back of house" promise, and distributors reach thousands of operators.

• Distributors move slowly

  Start with order-by-email or EDI files while the API deal is negotiated.

• Your websiteHomepage: headline "The autonomous back of house for restaurants." Four modules (Inventory, Labor, Finance, Intelligence): voice and vision counts, an AI assistant, autopilot agents with human approval, demand forecasting, agentic ordering, prep and scheduling. Eight customer logos including A&W Restaurants, The Kati Roll Company, Corgi Cafe, Ess-a-Bagel and Kitchen & Coffee. Integrations shown as logos only: Toast, Square, Clover, 7shifts, Sling, HotSchedules, ezCater, OpenTable, Resy, SevenRooms, QuickBooks, Xero, Restaurant365, Gusto, ADP. No pricing; every main call to action is "Get a demo".Source
• Competitor researchMarginEdge: "AI runs the numbers. You run the restaurant.", a published $350 per month per location, "Trusted by 12,000+ restaurants", distributor integrations such as US Foods, Gordon Food Service and Shamrock, setup within a few weeks. Restaurant365: 52,000+ restaurants, R365 AI with AI Scheduling in early access, and a mobile app in English and Spanish.See the research Source
• HiringFounding Engineer (New York, in office, $150K-$225K, 1%-3% equity): owns the agent harness, including human approval for purchasing decisions; builds POS, accounting and payroll integrations; makes agents reliable through evals and failure recovery.Source

Add a short Canada page (local POS and payroll tools, CAD pricing if published, Canadian proof) and decide whether Europe is a real market now or a few early accounts. If real, pick one country, add its integrations and a local case study; if not, say so and focus.

Making the existing international demand visible either opens cheap expansion or frees the team to focus.

• Spreading a team of three too thin

  Only do Canada unless Europe already has repeatable demand.

• HiringFounding Ops (New York, in office, $100K-$150K, 0.20%-0.75% equity): owns deployments from contract signing to go-live, is the primary customer contact, builds repeatable playbooks and supports sales through calls, pilots and multi-unit rollouts. The post says Marble signed 150+ restaurants in under four months across the US, Canada and Europe, growing 25% week over week for three months, with a team of 3.Source
• Your websiteReferral page: $500 on a prepaid card or an Oura Ring 5 ("About $569 in value") per referred restaurant, earned after 30 days of active deployment; rewards ship in the US; referrers enter a name and email to get a link; the homepage pitches it to non-operators.Source
• Competitor researchNory calls itself "the agentic AI restaurant operating system", also claims 97% demand forecast accuracy, has an Ordering Assistant in beta and an ROI calculator, and serves independent, franchise, multi-location and enterprise groups. Crunchtime cites 850+ restaurant brands and 150,000+ locations. MarketMan publishes pricing ($249, $299, from $449 a month) for 15,000+ restaurants.See the research Source

Support counts spoken in Spanish (and mixed English-Spanish item names) with the same item mapping, plus a Spanish app interface for counters. Mention it on the segment pages and demo, and track count completion by language.

It removes a barrier for the people who actually do the counts, which drives daily usage and the accuracy behind every forecast.

• Item mapping errors

  Confirm low-confidence items on screen before saving.

• Your websiteHomepage: headline "The autonomous back of house for restaurants." Four modules (Inventory, Labor, Finance, Intelligence): voice and vision counts, an AI assistant, autopilot agents with human approval, demand forecasting, agentic ordering, prep and scheduling. Eight customer logos including A&W Restaurants, The Kati Roll Company, Corgi Cafe, Ess-a-Bagel and Kitchen & Coffee. Integrations shown as logos only: Toast, Square, Clover, 7shifts, Sling, HotSchedules, ezCater, OpenTable, Resy, SevenRooms, QuickBooks, Xero, Restaurant365, Gusto, ADP. No pricing; every main call to action is "Get a demo".Source
• Competitor researchMarginEdge: "AI runs the numbers. You run the restaurant.", a published $350 per month per location, "Trusted by 12,000+ restaurants", distributor integrations such as US Foods, Gordon Food Service and Shamrock, setup within a few weeks. Restaurant365: 52,000+ restaurants, R365 AI with AI Scheduling in early access, and a mobile app in English and Spanish.See the research Source