# Response to Pitch Comments (28 Sept 2026) Responds to the comments in `2026_09_28_CommentsOnPItch.pdf` on [[Augmented Traveler Nantucket — Studio Pitch]]. Nearly every comment is right and should change the pitch. Answers below follow the order of the comments; open decisions are gathered at the end. ## The five things that matter most 1. **The money number in the pitch is too low.** Our own year-one budget (studio plus Mitchell) comes to $665k–$1.28M on a lean team, before founder pay. The pitch says $230k–$400k. We have to fix this before showing anyone a raise figure. 2. **Revenue is the weakest part of the pitch.** Island ticketing alone will not pay back the studio. The investor story has to be licensing characters and the engine to venues off the island, with Nantucket as the proof. 3. **An MVVP is the right first ask.** A 3–4 minute pilot with real visitor testing, roughly $60k–$120k, turns a $1M+ ask into a small, measurable first step. 4. **The first production should be chosen on evidence, not only on sources.** Mitchell is the safest build; whaling is the island's biggest draw. A cheap visitor concept test can settle it. 5. **Start with phone AR and kiosks; push headset rooms to Season 2.** Headsets add staff, hardware and setup cost before we know what visitors will pay for. ## 0. Summary **Three dimensions and presence.** Agreed. "Life-size" alone can read as video. Proposed opening line for the studio paragraph: > Augmented Traveler creates three-dimensional, interactive stories. It is like going to the theater and being allowed to walk up onto the stage: the characters share your space and your time, and sometimes they talk with you. We should also say once, plainly, what each surface does in 3D (see the walk-around answer in section 2). **Is Maria Mitchell the best first production?** Fair challenge; see sections 6–7 below. Short answer: she is the best first *character* (strongest sources, simplest ethics, one person to build). She may not be the best first *product* for paying audiences. The summary should say "proposed first production" until we decide. **Are setup and production costs additive?** Yes. $80k–$150k setup plus $150k–$250k for Mitchell is the $230k–$400k "indicative year one." But that total omits realistic personnel cost. Our detailed budget ([[Year One Budget — Combined Summary]]) puts year one at $665k–$1.28M (lean team) or $1.08M–$1.79M (full-time team), both before founder pay. See section 9/11. **MVVP (minimum valuable viable product).** Strongly agree. Suggested shape: - **What it is:** a 3–4 minute authored scene with one character, on one kiosk, plus one phone "postcard" on the street. A short list of authored answers to the top 10–20 questions, rather than open AI conversation. - **What it proves:** that the character is believable, that visitors stay, and what they would pay. - **Test plan:** one summer month (or a shoulder-season weekend run) at a partner venue. Measure dwell time, completion rate, repeat visits, a short exit survey, and a willingness-to-pay test (different prices shown to different groups). - **Rough cost:** $60k–$120k, including hardware that carries over into the full production. This is an estimate to be costed by the CTO. - **How to pitch it:** show the full numbers first, then offer the MVVP as the "test ride": investors cap their first risk at that amount and get real audience data before the larger commitment. ## 1–2. The idea and why now **"As one of several possible experiences, picture a summer evening…"** Agreed; adopt as written. **The comparables are not impressive.** You are right, and that is actually our argument. Proposed reframing of "Museums and destinations are already doing it": > Audiences already line up for historical characters, even in flat, limited forms. Nobody has yet put a full-body, three-dimensional, sourced character in the place where the person lived. That gap is the opportunity. How each example fits that framing: | Example | What it really is | Keep it? | | --- | --- | --- | | *Bonjour Vincent* (Musée d'Orsay) | Real-time Van Gogh on a screen, built from his letters | Keep as the closest comparable, and say plainly we aim to exceed it on believability. Your note ("too much artifact, too little character") is the bar we set ourselves. | | *Dalí Lives* (Dalí Museum, 2019) | AI-generated video of Dalí on screens: 2D, not interactive in 3D. Not the same as *Dalí Alive*, the museum's big-screen projection show. | Keep only as proof of demand (attendance), not as technology. | | *Dimensions in Testimony* (USC Shoah Foundation) | Recorded video answers from living survivors | Agreed, not our product. Move it out of "Why now" and cite it only in our standards, as the model for honesty about sources. | | *An Audience with a Queen* (Leeds Castle, 2026) | Historian-designed Eleanor of Castile, three languages | Keep; closest to our heritage-site model. | See also [[Case Study — Jumbo Mana Bonjour Vincent]] and [[Case Study — Dimensions in Testimony and StoryFile]]. **Can we walk around the character? Occlusion?** It depends on the surface, and the pitch should say so: - **Phone AR:** yes. The character stands in the real street; walk around her and see her from the side or back. Recent phones can hide her behind real people and objects ("occlusion"); it works best on iPhone Pro models with depth sensors and is rougher on older phones. - **Headset:** yes, fully. You are inside the room with her. (Our standards default to seated VR for accessibility, so walking around is optional.) - **Kiosk:** no. It is a window: the character is rendered in 3D but you see her from the front. A holographic-style display (such as Looking Glass) adds some depth as you move, at a smaller size. **"Kiosk is also the capture booth for live events": please explain.** Live Link Face is a free iPhone app that reads a performer's facial expressions through the phone's camera and streams them live into the 3D character. At a special event, a scholar or actor sits out of view with an iPhone pointed at their face and a microphone. The audience sees Maria Mitchell on the life-size screen, moving her face and speaking with the performer's live answers. It is a premium mode: a human answers instead of the AI, which is ideal for opening nights, donor events and school visits. The pitch should say this in one plain sentence or cut it. **Name the company that failed.** Yes: Soul Machines (New Zealand) entered receivership on 5 February 2026 after raising roughly US$120–135 million. See [[Digital Humans]]. **The three success factors leave out quality.** Agreed. Add a fourth: *the character has to be believable as a person*, which is exactly your criticism of the Van Gogh. **What is a "closed set of sources"?** It means the character may answer only from a fixed library of documents that the historian has approved: her letters, diaries and accounts by people who knew her. It does not draw on the open internet or the AI's general knowledge. If the answer is not in that library, she says she did not write about it. Proposed plain wording: "a fixed, historian-approved library of sources." ## 4–5. What the studio is, and our standards **How motion capture works, and why it matters.** Proposed one-paragraph explanation for the pitch: > An actor in costume-weight clothing performs the scene while cameras or a sensor suit record every movement of the body, hands and face. That recording is applied to the digital character's skeleton, so she moves with a real person's weight, timing and small gestures. Animating ten minutes of a speaking person by hand takes weeks and still looks stiff; captured movement is what makes a visitor forget they are looking at a computer model. We also capture "idle" takes (breathing, glancing, shifting weight) so the character looks alive while she listens. **Should a character sometimes be a companion?** Yes, as a *guide*, and the pitch should make the distinction. What our rules ([[Ground Rules — Historical Virtual Humans]]) forbid is emotional companionship: "I'm your friend," "I love you," keeping secrets, especially with children. A character who walks you through town and points out sites is a docent, which the rules allow. Rewrite the line as: "A character may guide you; it will never pose as your friend." On who guides the walk, two cautions and a suggestion: - **Ahab and Starbuck are fictional.** They work as a labeled literary scene at the Whaling Museum stop, but a fictional guide to real history mixes the two in visitors' minds. - **Obed Macy is a strong real candidate.** He kept journals from 1799 to 1844 and published the island's first history in 1835; our research ([[Nantucket Historical Figures — Opportunities]]) already flagged him as a natural town-walk narrator. - **Make the walk audio-first.** The guide talks in your earbuds as you walk and appears in phone AR only at key stops. That is far cheaper to build than a character visible the whole way, and much easier on phone batteries. **Putting a character through a reading list.** This is a good idea and fits how the AI mind already works. We do not "teach" the character by having it read; we give it a library it must look things up in before answering, and every answer is tied to a source. A syllabus would become part of the period-knowledge library: books and papers the person is documented as having read, or that an educated islander of that year could have read. Two rules apply: nothing published after the character's cutoff year, and the historian approves the list. The result is a character who can talk knowledgeably about her world, not only about herself. Voice and manner (how she phrases things) can additionally be tuned, but facts always come from the library. **Chautauqua format.** Agreed; adopt your description in the pitch: 1. An in-character autobiographical performance. 2. A question-and-answer session in character. 3. A final session where the scholar steps out of character to give objective historical analysis. In our version, step 3 can be a live scholar, a clearly labeled narrator, or wall text. ## 6–7. The slate and the first production **Add the popularity question.** Agreed. The six questions include "Is there an audience?" but it is the weakest one, and investors will weigh it first. Replace it with a stronger seventh-and-first question: *Will people pay for this, and how many?* Score it from evidence: NHA and MMA attendance by site, sales of comparable programs (NHA walking tours, MMA evening observatory sessions), and our own visitor test. **Drop the status column; rank provisionally.** Agreed. Proposed wording: "Ranked by our current assessment. The first production will depend on partner commitments and who funds it." **Mitchell or the returning whaleship?** An honest comparison: | | Maria Mitchell | Returning whaleship | Essex survivors | | --- | --- | --- | --- | | Audience pull | Families, schools, science fans; narrower | Whaling is the island's brand; broadest | High (*Moby-Dick*, *In the Heart of the Sea*) | | What visitors meet | One person who talks with them | A scene and a crowd; only a harbor clerk answers | Two to three people who disagree | | Build cost and risk | Lowest: one character, simple ethics | Higher: ship, wharf, cargo, crowd | Highest: trauma protocol, two corpora | | Lead partner | Maria Mitchell Association (smaller) | NHA Whaling Museum (largest venue) | NHA | | Proves the core product? | Yes: a believable person you can question | Partly: spectacle more than conversation | Yes, but a hard first build | One way to have both: **use Mitchell for the MVVP** (cheapest proof that the character engine works) and **keep the whaling story as the flagship candidate** for the full first production with NHA, if the visitor test favors it. The ship and wharf set is also reused by three other slate projects, so building it early is not wasted. See [[Project — Returning Whaleship]] and [[Tier 1 — Maria Mitchell]]. **Should investors rank the possibilities?** Inviting their view builds engagement, but presenting no ranking can read as indecision. Better: show our ranking with reasons, then say "a lead funder may sponsor a specific figure." **How much to rely on AI for market analysis.** Use AI for desk research, competitor scans and drafting surveys. Do not present AI output as market evidence; investors will discount it. The evidence should be primary: partner attendance figures, comparable ticket sales, and a visitor concept test. A cheap version: show 300–500 visitors one-page concept cards for four or five experiences and ask which they would buy and at what price (a few thousand dollars, mostly staff time). ## 8. Experiences across the island **Cost estimates for the five loops.** First-pass, order-of-magnitude figures for the CTO to confirm. They assume the phone app framework is built once (about $40k–$80k, part of studio setup) and that characters already built for other productions are reused. | Loop | Stops | What each stop is | Rough cost | | --- | --- | --- | --- | | Downtown core | 6 | Short character scenes plus audio | $40k–$90k | | Vestal Street | 4 | Mitchell postcards (reuses her character) | $20k–$50k | | New Guinea | 2–3 | Scenes, partner-led with MAAH | $20k–$45k plus partner costs | | Water and mills | 4 | Audio guide with a few AR moments | $25k–$60k | | Far island | 4–6 | Landscape and audio cards, no characters | $10k–$25k | All five loops together: roughly $115k–$270k on top of the app framework, built over two or three seasons. Per stop, a scene with a new character costs far more than one reusing an existing character, which is another argument for building the slate on shared characters. **Start with phone AR instead of headset rooms.** Agreed. Headset rooms need staff on duty, device cleaning and charging, onboarding for every visitor, and about $1k–$3.5k per headset. Phone AR uses the visitor's own device, and it leaves nothing physical in the historic district (a kiosk or sign there likely needs Historic District Commission approval; an AR scene does not). Proposed sequence: Season 1 = kiosk + phone AR; headset rooms move to Season 2, once visitor data shows demand. The [[Vestal House VR Room]] can still be built as a set, since the same 3D room also serves the kiosk background. ## 9 and 11. The money **Telescope the funding sections.** Agreed. Proposed structure: the summary gets one short paragraph (production cost range, the MVVP "test ride," the total raise). Everything else moves into a single section, "The money," in this order: total raise and runway, use of funds, revenue model, milestones. **Total raise and runway.** The pitch's $230k–$400k year one does not survive our own budget, which adds personnel at market rates: | Scenario | Year one (12 months) | 18-month runway | | --- | --- | --- | | Lean core + contractors (5 full-time) | $665k–$1.28M | about $0.95M–$1.78M | | Full-time studio (13 full-time) | $1.08M–$1.79M | not modeled | Both exclude founder pay, rights counsel, collections licensing, venue costs and marketing, so the real figure is higher. The 18-month figure adds two quarters at the lean team's year-end spending rate. We either raise the pitch number to this level or cut the first production's scope. A staged ask makes the larger number easier to hear: 1. **Concept test:** a few thousand dollars. Visitor survey on four or five experiences. 2. **MVVP:** $60k–$120k over three to four months. Pilot plus measured visitor testing. 3. **Seed round:** roughly $1M–$2M for 18 months. Studio, first full production, Season 1 opening. **A rigorous revenue workup.** Agreed this is the most important gap. Below is a first model for Season 1 (Mitchell kiosk plus downtown and Vestal phone loops). Every input is an assumption to replace with partner data and MVVP results; the point is to show investors the method and which numbers matter. | Revenue stream | How it is calculated | Low | Mid | High | | --- | --- | --- | --- | --- | | Evening kiosk sessions | sessions × guests × price × our 50% share | 40 × 12 × $20 = $4.8k | 70 × 18 × $25 = $15.8k | 100 × 25 × $30 = $37.5k | | Phone loop passes | buyers × price, less store or payment fees | 2,000 × $8 = $12k | 6,000 × $10 = $48k | 15,000 × $12 = $144k | | Venue commission (MMA/NHA) | negotiated fee | $0 | $50k | $150k | | Schools (classroom mode) | schools × annual fee | 5 × $500 = $2.5k | 15 × $1,000 = $15k | 30 × $1,500 = $45k | | Grants | awards received | $25k | $75k | $200k | | **Season 1 total** | | **about $44k** | **about $204k** | **about $577k** | What the model tells us: - **The island alone will not repay the studio.** Even the high case is below the lean year-one cost. Investors will see this, so we should say it first. - **The swing factor is phone loop sales.** 15,000 buyers is a small share of a summer in which the island's peak-day population now exceeds 60,000, but we have no conversion data yet. The MVVP must measure it. - **The recurring income story is licensing.** Once a character, room set and program exist, licensing them to off-island venues (whaling museums, planetariums, abolition sites) costs little to deliver. Illustratively, ten venues paying $50k a year is $500k of recurring revenue on characters already paid for. We need two or three letters of interest from such venues to make this credible. - **Next step:** get attendance figures from the NHA annual report and MMA, and price data for comparable island programs, then rebuild this table with real inputs. ## Miscellaneous **Island organizations to contact.** A starter list for Nancy Martin to correct, extend and prioritize: | Organization | Why it matters to us | | --- | --- | | Nantucket Historical Association | Manuscripts, Whaling Museum, most venues | | Maria Mitchell Association | Mitchell house, papers, instruments | | Nantucket Atheneum | 1841 Douglass hall; Mitchell as librarian | | Museum of African American History (African Meeting House) | Required partner for New Guinea and Absalom Boston | | Egan Maritime Institute (Shipwreck & Lifesaving Museum) | Far-island loop; maritime stories | | Nantucket Preservation Trust | Historic buildings; walking-tour audiences | | ReMain Nantucket | Downtown vitality; fits the downtown loop | | Historic District Commission (Town) | Approval for any physical kiosk or sign | | Nantucket Chamber of Commerce and Town Visitor Services | Distribution of loop maps to arriving visitors | | Nantucket Cultural District | Cross-promotion among cultural venues | | Nantucket Public Schools | Classroom mode, year-round audience | | Theater Workshop of Nantucket | *Spirits Within Us*; actors; collaboration | | Steamship Authority and Hy-Line | Reaching visitors on the ferry | | Aquinnah and Mashpee Wampanoag cultural offices (off-island) | Required before any Wampanoag content | **Wendy Schmidt and ReMain Nantucket.** The risk is real but manageable. Our protection is not secrecy (investors rarely sign NDAs); it is partnerships and execution. Sign the MMA or NHA agreement before approaching large local philanthropists, so the institutions are already committed to us. Share the vision deck, not the production methods and budget detail. Seek a warm introduction, and approach ReMain as a potential partner or funder for the downtown loop rather than as a threat; a funder who backs us is the best defense against one who competes. **Sound: are Bluetooth earbuds practical?** Yes, for the street. Most visitors already carry them, and AirPods-class earbuds can make the voice seem to come from where the character stands. Three cautions: Bluetooth adds a delay that can make lip-sync look off, so street scenes should favor narration and medium shots over close-up talking; the phone speaker must work as a fallback; and captions stay on. At kiosks, use directional speakers or a sound dome so the gallery is not flooded. Headsets have built-in audio. **Phone updates across manufacturers.** This is manageable with a stated policy rather than a solution: - **Support window:** phones from roughly the last five years. iPhones are simpler (Apple updates all models at once); Android variety is the real work. - **One codebase:** build in an engine that handles both Apple's and Google's AR systems, so we fix once, not twice. - **Fallbacks:** a no-install version opened from a QR code, and an audio-and-video version for phones that cannot run AR. Every visitor gets something. - **Test lab:** eight to ten representative phones, retested before each season and after major OS releases; the cost sits in the annual maintenance line. - **Loaner devices:** a few preloaded tablets at Visitor Services or partner front desks. - **Avoid single-vendor dependence:** a popular web-AR platform, 8th Wall, shut down in February 2026, stranding projects built on it. ## Decisions to discuss - [ ] Raise the pitch's year-one figure to the budget's level, or cut the first production's scope? - [ ] Approve the staged ask: concept test, MVVP ($60k–$120k), then a $1M–$2M seed round? - [ ] Run the visitor concept test before naming the first production? - [ ] Mitchell for the MVVP and whaling as flagship candidate, or one figure for both? - [ ] Keep the Live Link Face live-performance mode in the pitch, or cut it? - [ ] Audio-first walking guide: Obed Macy, or another real Nantucketer? - [ ] Move headset rooms to Season 2? - [ ] Which off-island venues could give letters of interest for licensing? - [ ] Timing and route of any approach to ReMain Nantucket? Once these are settled, the pitch can be revised in one pass. ## Sources - [[Year One Budget — Combined Summary]] - [Nantucket peak-day population, CAI / Nantucket Data Platform](https://www.capeandislands.org/local-news/2022-07-27/new-data-shows-nantucket-is-on-track-to-host-100-000-visitors-on-a-peak-summer-day-by-2030) - [Nantucket summer 2025 tourism roughly flat, Nantucket Current](https://nantucketcurrent.com/news/data-indicate-tourism-to-nantucket-was-roughly-flat-during-summer-2025) - [Dalí Lives, Salvador Dalí Museum](https://thedali.org/exhibit/dali-lives/) - Soul Machines receivership: see [[Citations — Digital Humans]]