Content Production — scripts, outlines, and the tooling reality

Owner: Krishna Malyala · Date: 2026-06-25 · Status: DRAFT — nothing publishes without Krishna Feeds: gtm/CAMPAIGN-PLAN.md (the 8-week calendar references these by number) · Numbers: CAMPAIGN-PLAN §0 (verified only)

One content engine feeds every platform. Each long-form YouTube spawns 3–4 Shorts; each Short cross-posts to TikTok / Reels / Facebook; each data story becomes a LinkedIn carousel; the blog posts are the SEO/GEO backbone and the email-gate destinations. The funnel tag on each piece tells you which lane (A=investor, B=acquirer, C=consumer) it serves.

0. TOOLING REALITY — what exists on the fleet today (investigated 2026-06-25)

Before writing a single avatar script, here's what's actually installed vs. what needs setup. Bottom line: we have zero avatar/TTS capability installed today, but the gap is a 2–3 hour, ~$30–110/mo fix.

Recommended minimal stack to start (Tier-1 budget, ~$50–110/mo): 1. HeyGen creator (~$30–90/mo) — the 6 avatar Shorts. A consistent branded avatar = the channel's face without filming Krishna. 2. ElevenLabs creator (~$22/mo) or Piper ($0) — voiceover. ElevenLabs if we want a cloned Krishna voice for authenticity; Piper if budget = $0. 3. apt install ffmpeg + Canva Pro ($13) — mux, caption, thumbnails, carousels. 4. Remotion bootstrap (free, Node already here) — animated data Shorts + the "Losing Ground 2026" chart animations. 5. claudetube (installed) — reference-pull + auto-transcribe/caption.

Production flow: script (below) → ElevenLabs/Piper voiceover → HeyGen avatar OR Remotion data-animation → ffmpeg caption+mux → Canva thumbnail → schedule via Buffer → host MP4 on videolens/geo. First avatar Short shippable within ~2–3 hours of subscribing.

1. Short-form avatar scripts (YouTube Shorts / TikTok / Reels)

Format for each: Hook (0–3s) · Script (~30–45s) · On-screen text · CTA · Funnel tag · Pillar. All claims restricted to CAMPAIGN-PLAN §0 verified numbers. Avatar speaks; on-screen text reinforces; B-roll = the live map / receipt.

Short #1 — "The $2,302 hidden commute tax" · Pillar P1 · Lane A+C

Hook (0–3s): "Your map app is hiding a $2,302-a-month number from you." Script (~35s):

"Every map will tell you Summit, New Jersey to Manhattan is about an hour — whether you drive or take the train. What none of them tell you? Driving it costs two thousand three hundred dollars a month more than the train. That's tolls, that's parking, that's congestion pricing, that's five-dollar gas — none of it shows up when you only see the time. We built the map that shows the true cost of any trip, by every mode. Time is a solved problem. Cost is the one wrecking budgets — and nobody was pricing it. Until now."

On-screen text: 1 hour either way → DRIVE: $2,302/mo MORE → tolls · parking · CRZ · $5 gas → commutingcost.com CTA: "See what your commute really costs — link in bio." Funnel: drives to the free map (Lane C capture); the "nobody prices it" line plants the moat for Lane A.

Short #2 — "Find the home that works for BOTH of you" · Pillar P2 · Lane A+B+C (the WOW)

Hook (0–3s): "Zillow can't answer this. Google can't. We can." Script (~40s):

"Here's a question no map app can answer: where should a couple live — when one person works in Midtown and the other in Jersey City? Every tool answers for one person, one commute. But a household has two. So we built it: enter both jobs, and the map shades every home green where it works — in time AND money — for both of you. Not 'close-ish to one job.' Affordable, for both commutes, across nine modes. It's like a home-match index, but it actually knows what the trip costs. Nobody else does this. Drop your two job ZIPs and watch."

On-screen text: 2 jobs. 1 home. → Midtown + Jersey City → map shading green → works for BOTH ✓ → time + cost · 9 modes CTA: "Try the 2-commuter finder — link in bio." (NOTE: only run this CTA once the /map '2 commuters' mode is live — see CAMPAIGN-PLAN §8 risk.) Funnel: the flagship differentiator; Lane B's capability-they-lack proof, Lane C's share-bait.

Short #3 — "How we reprice a whole city in seconds" · Pillar P3 · Lane A

Hook (0–3s): "This used to take 10 days. Now it takes 21 seconds." Script (~35s):

"When congestion pricing in Manhattan jumped to nine dollars, most cost-of-living tools were instantly wrong — and stayed wrong for a year, because they update once a year. Ours repriced the entire metro in one pass. Here's why: the old way of computing every commute cost in a region took about ten days on a normal computer. We rebuilt it to run on a single twelve-hundred-dollar graphics card — and now it takes twenty-one seconds. That's not a flex. It's the reason our cost map is live and theirs is a stale PDF. Fares change, tolls change, schedules change — we keep up. Annual indices can't."

On-screen text: 10 DAYS → 21 SECONDS → 1× $1,200 GPU → CRZ → $9? repriced in 1 pass → live, not annual CTA: "How it's built — full breakdown on YouTube." Funnel: Lane A credibility (freshness-as-moat); note P3 is framed as why the cost layer is live, never as the headline.

Short #4 — "Housing + Transportation + Utilities" · Pillar P4 · Lane A+C

Hook (0–3s): "You're budgeting for a home with one-third of the real number." Script (~35s):

"Everyone knows to ask: can I afford the rent? Smarter people add the commute — that's what the Center for Neighborhood Technology's famous H+T index did: Housing plus Transportation. But there's a third number families actually feel, especially now: utilities. Gas AND power are both climbing. So we built H + T + U — Housing, Transportation, Utilities — in real dollars, at the address level, for any home. The 'affordable' neighborhood that's a brutal commute and a $400 electric bill? Not affordable. Now you can see the whole number before you sign."

On-screen text: H + T = old way → + U (utilities) → H + T + U = real cost → address-level $ CTA: "What's the real cost of your neighborhood? Link in bio." Funnel: Lane A (the validated-index story), Lane C (relatable); CNT attributed on-screen + in caption.

Short #5 — "Your city is losing ground" · Pillar P5 · Lane B+C+press (REPORT LAUNCH, 7/28)

Hook (0–3s): "In 2010, a landmark study said working families were losing ground. We re-ran it for 2026." Script (~40s):

"Fifteen years ago, the Center for Neighborhood Technology published 'Losing Ground' — proof that working families were being crushed by the combined cost of housing and getting to work. We just re-ran that study with 2026 data, across 32 metros, in a five-dollar-gas economy — and added what they couldn't: live, address-level commute costs by every mode, plus utilities. The findings are worse than you'd guess, and they're different in your city. We're releasing the whole thing. This is the number the affordability conversation has been missing."

On-screen text: LOSING GROUND → 2010 → 2026 → 32 metros · $5 gas → the full report ↓ CTA: "Read 'Losing Ground 2026' — free, link in bio." (email-gate) Funnel: the campaign centerpiece; Lane C/press reach → email capture → Lane A/B mining. Make 3 city-specific versions for the metros most affected.

Short #6 — "Built with a $1,200 GPU and a lot of stubbornness" · Pillar P3 · Lane A (founder/build-in-public)

Hook (0–3s): "One person. One $1,200 graphics card. 32 cities." Script (~35s):

"People ask how a tiny team out-built the cost layer that Google, Zillow, and Walk Score never shipped. Honest answer: the math is public — anyone can read the routing paper. The hard part nobody wants to do is the data: cleaning 723 transit fare feeds, tolls across 36 states, parking garages in 49 metros, bike-share pricing for 194 systems — and a GPU pipeline that reprices it all in seconds. That's years of unglamorous work plus one cheap graphics card. The result: the true cost of any trip, any mode, 32 metros. The moat isn't the algorithm. It's the data and the speed."

On-screen text: 723 fare feeds → tolls · 36 states → 1,515 garages → 194 bike systems → 32 metros · 1 GPU CTA: "Investors / builders — DMs open." (LinkedIn-native version) Funnel: Lane A — the "founder ships + real moat" signal; also seeds Lane B's "capability you'd spend years rebuilding."

(Optional) Short #7 — "What $5 gas actually costs your family" · Pillar P1 · Lane C

Hook: "$5 gas isn't a headline. It's $1,924 a month you can't see." Script (~30s): quick punchy version of the rail-vs-drive saving, generalized: "Across 30 tri-state towns, taking the train instead of driving saves the average family nineteen hundred dollars a month. Rail won in all thirty. Most people have no idea — because no map shows the cost, only the clock. Here's the one that shows both." On-screen: $1,924/mo avg saving → rail cheaper 30/30 → you just can't see it CTA: "Check your town — link in bio." Funnel: Lane C reach.

(Optional) Short #8 — "The 2×2 nobody fills" · Pillar P1+P2 · Lane A+B

Hook: "Here's the entire trip-intelligence market in one square." Script (~35s): animated 2×2 (Remotion). "Trip tools have two axes. Time, or time plus cost. One person, or the whole household. Google Maps and TravelTime nailed exactly one box — time, one person — and stopped. The other three boxes? Cost. Multimodal. Multiuser. Empty. And they're the exact dimensions that move a family's budget. That empty space is the company." On-screen: the 2×2 filling in, three quadrants lighting up with our logo. CTA: "The full thesis — YouTube." Funnel: Lane A/B — the wedge, visualized.

2. Long-form YouTube outlines

Format: title · length · audience · the binge-arc · sections · CTA.

Long-form #1 — "Why no map app can tell two people where to live" · ~8 min · Lane A+B

The investor/acquirer binge artifact — the multiuser wow, explained. 1. Cold open (0:00–0:45): the two-job problem — Midtown + Jersey City. Why "20 min to one job" is the wrong question for a household. 2. The market gap (0:45–2:30): the 2×2. Time vs. cost, one-person vs. household. Google/TravelTime own one box. The other three are the opportunity. Name them as the incumbents we extend, not enemies. 3. The cost stack (2:30–4:30): what we price that nobody does — fares (723 feeds), tolls (36 states), CBD parking (1,515 garages), CRZ, fuel, GBFS (194 systems). The receipt for one trip. 4. The multiuser math (4:30–6:30): intersect two time+cost isochrones → the home that works for both. Screen-record the live demo. Why it's a different product shape, not a feature toggle (the thing incumbents can't express). 5. Why it's live, not annual (6:30–7:30): 10 days → 21 s on a $1,200 GPU; the freshness moat (CRZ → $9 repriced in one pass). 6. Close (7:30–8:00): where it applies beyond real estate — jobs, events, relocation. CTA: "Try it / investors, link below." CTA: map link + gated investor portal.

Long-form #2 — "H+T+U: the affordability number that's been missing a leg" · ~7 min · Lane A+C

The validated-index credibility piece. 1. The history (0:00–1:30): CNT's H+T index — adding Transportation to Housing changed affordability thinking. Credit it fully. 2. The missing leg (1:30–3:00): Utilities. Why now (gas + power both climbing). Address-level utility rates (3,230 utilities, EIA-861). 3. Putting it together (3:00–4:30): H + T + U in dollars and % of income; HUD-AMI vs. CBSA median (+47% — the denominator matters). 4. Did we get it right? (4:30–6:00): validation — r = 0.90 vs CNT H+T, 4/5 OD pairs, 12/12 walk legs. Show the rigor (investors love this). 5. The Summit example (6:00–6:45): $3,178/mo all-in; rail saves $2,302/mo. The receipt. 6. Close (6:45–7:00): CTA — check your county / methodology link. CTA: TLCcosts pages + public methodology.

Long-form #3 — "Losing Ground 2026: we re-ran the landmark affordability study" · ~10 min · Lane B+C+press (CENTERPIECE, 7/28)

The definitive report video — the artifact press embeds. 1. The 2010 original (0:00–1:30): what CNT's "Losing Ground" found, why it mattered, attribute fully. 2. Why re-run it now (1:30–3:00): $5 gas, congestion pricing, post-pandemic relocation, utilities climbing. The 2010 method can't see live cost. 3. What we added (3:00–4:30): live address-level commute cost, 9 modes, 32 metros, utilities. The methodology, honestly (modeled drive time flagged; CNT vintage caveat). 4. The findings (4:30–8:00): the headline finding + the 5–6 charts that matter. City-by-city. The metros losing the most ground. (Fill with the real report once produced.) 5. What it means (8:00–9:30): for families, for housing policy, for where the affordability conversation goes. 6. Close (9:30–10:00): download the full report (gated). CTA for press: "data + methodology available." CTA: gated report download; press contact.

Long-form #4 (optional) — "The 4 things every map app is still missing" · ~6 min · Lane B

The acquirer-aimed capability-gap piece (run Week 7). Walks cost / multimodal / multiuser / instant-national-reprice as the four gaps, each with the live proof, framed as "the next layer the category's waiting for." Ends without an "acquire us" — just "this is the upgrade." (See ACQUISITION-POSITIONING §2 for the capability framing.)

3. Blog posts (SEO/GEO backbone + email-gate destinations)

5 titles + outlines; #1 fully drafted (~800 words).

1. "The $2,302 Question: Why Your Map App Hides the Real Cost of Your Commute" — fully drafted below. Lane C+A. SEO: "real cost of commuting," "drive vs train cost."
2. "H+T+U: The Affordability Number That's Been Missing a Leg" — outline: CNT history → utilities gap → the validated index → HUD-AMI denominator → try-it CTA. Lane A+C. SEO: "true cost of living by neighborhood," "housing transportation utilities affordability."
3. "Two Jobs, One Home: How to Find Where a Couple Should Actually Live" — outline: the dual-commute problem → why single-origin tools fail → the multiuser finder → how to use it → cases (rail+bike split). Lane C+B. SEO: "where should a couple live two jobs," "dual commute home search."
4. "We Repriced 32 Cities in Seconds. Here's Why Your Cost-of-Living Data Is Always Stale." — outline: the freshness problem (annual indices) → CRZ-to-$9 example → 10 days→21s → what live cost data unlocks. Lane A. SEO: "cost of living data accuracy," "live commute cost."
5. "Losing Ground in 2026: What a $5-Gas Economy Did to America's Commuter Towns" — outline: the report's headline findings, blog-form, as the launch companion + lead magnet. Lane B+C+press. SEO: "losing ground report 2026," "transportation cost burden by city."

Blog #1 — FULLY DRAFTED (~800 words)

The $2,302 Question: Why Your Map App Hides the Real Cost of Your Commute

Open Google Maps and ask it how to get from Summit, New Jersey to Midtown Manhattan. It will give you a clean answer: about an hour. Drive or take the train, it's roughly the same clock. Pick whichever you like.

That answer is a trap — because it's only half the picture, and it's the wrong half.

Here's the part no map app shows you: driving that commute costs about $2,302 a month more than taking the train. Not a little more. Twenty-three hundred dollars a month — the difference between a comfortable budget and a tight one. It's there in the tolls, the Manhattan parking, the $9 congestion charge, and gas that's hovering near $5 a gallon. But you'll never see it, because every mapping tool ever built optimizes for one thing: time. They answer "how long." They have never answered "how much."

For decades, that was fine. Time was the scarce resource, and gas was cheap enough that the cost of a commute hid in the noise of a household budget. That era is over. With gas near $5, transit fares rising, tolls climbing, and congestion charges spreading, the cost of getting to work, school, and life has become the thing that actually hurts. People are making the biggest financial decisions of their lives — where to live, where to work — using a map that shows them only half the equation.

We built the other half.

commutingcost.com computes the true, door-to-door dollar cost of any trip — not just the time. Pick a starting point, pick a destination, and we show you what it costs to get there by every mode: walk, bike, scooter, drive, transit, and the chained options most tools ignore entirely — Park & Ride, Kiss & Ride, and Uber-plus-rail. Each mode comes with a real receipt: the fare, the tolls, the parking, the congestion charge, the fuel. Posted prices, not guesses. Across 32 metros in the US and Canada.

The numbers surprise almost everyone. Across 30 tri-state commuter towns, taking the train instead of driving saves the average household $1,924 a month — and rail came out cheaper in all 30. People are quite literally driving past thousands of dollars a month in savings, because no one ever showed them the price tag.

There's a reason this layer didn't exist before, and it's not that nobody thought of it. It's that pricing a commute is genuinely hard. It means assembling and constantly cleaning a national database of posted prices — transit fares from 723 different agency feeds, tolls across 36 states, parking rates from over 1,500 downtown garages, bike-share pricing from 194 systems, fuel prices down to the ZIP code. Then you have to recompute all of it every time a fare changes, a toll goes up, or a transit board votes. When Manhattan's congestion charge moved to $9, the static cost-of-living indices everyone cites were instantly wrong — and stayed wrong, because they update once a year. Ours repriced the entire region in a single pass, in seconds.

That speed isn't a gimmick. It's the reason the cost is live and not a stale annual PDF. We rebuilt a computation that used to take ten days on a normal computer so it runs in about 21 seconds on a single graphics card. Fares change, schedules change, tolls change — and the map keeps up.

And there's one thing we can do that no map app can: solve for a household. A family isn't one commuter. When one partner works in Midtown and the other in Jersey City, "20 minutes from one job" is the wrong question. So you can enter both jobs, and we'll shade every home by whether it works — in time and money — for both commutes at once. It's like a home-match tool that actually knows what the trip costs. No other product can express it, because every other product was built for one person.

This isn't only about real estate. The true cost of a trip changes the math on where to take a job, whether a concert across town is worth it, what a relocation actually costs. Anywhere a trip has a price, the price should be visible.

Google Maps is brilliant at telling you a commute takes an hour. The question that's wrecking budgets in 2026 isn't how long the trip is. It's how much it costs — by every mode, for everyone who has to take it. That's the question we answer.

See what your own commute really costs — and what it would cost from anywhere you're thinking of living — at commutingcost.com.

(Methodology and sources are public. Cost figures use posted prices; drive times are modeled. Benchmarked against the Center for Neighborhood Technology's H+T® Index at r = 0.90; H+T® Index © CNT, used for comparison.)

4. Podcast episode outlines

3 episodes — pitchable to existing shows OR a founder-hosted mini-series. Each tied to the funnel.

Podcast #1 — "The number maps don't show you" · Lane A+C · target: personal-finance & proptech shows

• Arc: the $5-gas affordability crisis → why time-only maps miss the budget pain → the true-cost thesis → the Summit $2,302 story → where it applies beyond commuting.
• Pitch to host: "I can give your audience a concrete, surprising number about their own commute — and the bigger story about why no one priced it until now."
• Funnel: drives to the free map (Lane C); the moat + market-timing narrative seeds Lane A listeners (proptech-show audiences skew investor/operator).
• Pitch targets: personal-finance pods (FIRE/Money), proptech pods (e.g., proptech-VC-hosted shows), urbanist pods.

Podcast #2 — "Building a national geospatial product with one GPU and AI" · Lane A · target: tech/build-in-public & AI-ops shows

• Arc: the build story — replacing a 10-day CPU job with a 21-second GPU kernel → the data grind (723 feeds, the orphan-service-id bug class) → multi-region engine → how AI-assisted solo-building made 32 metros feasible.
• Pitch to host: "A real, shipped, national geospatial system built mostly solo with AI — the GPU/RAPTOR engineering and the data-cleaning war stories."
• Funnel: pure Lane A — the "founder ships + technical moat is real" signal that investors and acqui-hire-minded acquirers (Lane B) both weight heavily.

Podcast #3 — "Losing Ground in 2026: the affordability story everyone's missing" · Lane B+C+press · target: urbanism/policy & housing shows (CENTERPIECE TIE-IN, late July)

• Arc: CNT's original "Losing Ground" → why re-run it → the 2026 findings → policy implications (congestion pricing, transit funding, where families can afford to live).
• Pitch to host: "We re-ran a landmark affordability study with 2026 data across 32 metros — here are the findings, exclusively, ahead of the full report." (Time to the 7/28 launch; coordinate with PR-OUTREACH embargo.)
• Funnel: press/authority (Lane B + C), report email capture; gives the report a spoken artifact for the launch wave.

5. Production checklist (do this once, then it's a factory)

• Subscribe HeyGen (or D-ID) + ElevenLabs (or install Piper). Create one consistent branded avatar + voice.
• apt install ffmpeg on the production host; bootstrap Remotion (npx create-video@latest).
• Build the 6 core avatar Shorts (#1–#6) first — they carry the calendar.
• Render Long-form #1 (multiuser) and #3 (Losing Ground) as the binge artifacts.
• Draft & schedule the LinkedIn carousels from the data stories (Canva).
• Stand up the email-gate on the "Losing Ground 2026" report (the Lane-C→A/B bridge).
• Build the multiuser /map demo OR record an honest concept demo before Short #2's CTA goes live (CAMPAIGN-PLAN §8 risk).
• Host final MP4s + report HTML on geo/videolens; schedule via Buffer/Publer per the 8-week calendar.