Easy Budget

Easy Budget uses AI to replace producing time. It lets a commercial producer build, award and track an AICP budget through the whole life of a job — bid to wrap to final cost report — cutting the work from about five hours to about thirty minutes.

Category — deliberately two-sided. Capturing existing spend: AI-disrupted production budgeting, taking the Hot Budget seat producers already pay for and delivering it in the cloud instead of a macro-locked single-seat workbook. Creating new demand: agencies and brands scoping work can talk to a trained agent to project costs and budget outbound RFPs — buyers who do not budget today, because existing tools require a producer to drive them.

Customer sequence. Customer 0 is the freelance commercial line producer. The ultimate customers are agencies and production companies. Long term: anyone with complex production to budget and spend to manage.

ICP. Commercial line producers — the person who builds and owns the bid. National spots, socials, photoshoots and web content; explicitly not indie or documentary. Early users budget up to ~$1.5M a job. They range from freelancers bidding regionally in Dallas or Nashville to producers running jobs for Biscuit and Prettybird.

Trigger. The incumbent fails mid-bid. Hot Budget crashes; the segment is already migrating, which is why Saturation.io is gaining traction. We are not creating dissatisfaction — we are redirecting it.

Knowledge edge. Twenty years as a producer of broadcast commercials, socials and features, and a veteran partner at multiple regional and national commercial production companies. I have been the ICP for two decades. I also know the industry's referral topology — which producers get hired by which companies — so we target the node, not the leaf. And the incumbent's failure modes are documented, not remembered: a full Hot Budget teardown (May 2026) with a feature inventory, a calc-logic reverse-engineering of fringe base, markup stacking and overtime tiers, and a scored AI-opportunity backlog.

Distribution edge. Founder-level relationships at Softspot, Smuggler, Directors Bureau and post house Cabin Edit. 100+ producers in the personal network. The producers we hire today to make work for brands including Coca-Cola and Adobe. Beta users include Maddie Clarke (founder, YadaYada) and Micah Ross (founder, Arden Productions).

Why doesn't this exist already? Two reasons that compound.

01 · The economics never justified it. Commercial production and post is a $5B+ US industry — large enough to matter to an operator, far too small to interest a company optimising for venture-scale returns. So the market leader is a macro-locked Excel workbook, and nobody rebuilt it.

02 · Nobody else sees the full opportunity. The prize is not budgeting software. It is replacing producer labour and getting paid for the result — then owning the relationship with the deciders who run a very large number of very small companies, and selling them payroll and insurance. Priced as a budgeting tool the market looks too small to bother with. Priced as labour replacement plus a services funnel, it is a real business.

The wedge against Saturation is focus. Easy Budget does just commercials. That constraint is the advantage: one clean tool built perfectly for one job, against a competitor splitting attention three ways. Switching cost is already lowered — Saturation JSON import is built, so their users can leave with their data.

What happens to the gap. It closes if an incumbent ships credible AI, on a clock of roughly 18–36 months. The defence is speed and compounding data: AI-first architecture is a head start rather than a feature; real bidding data makes the agent progressively better at bidding, running and actualising; and founder-level distribution locks the corridor before a competitor reaches it. This is a speed bet and we say so.

Why now, two legs. The AI leg: an agent can now interview a producer and populate a structured AICP budget, and read a messy existing one. That was not possible on prior models at prior prices. The stronger leg is non-AI: the incumbent is actively failing and the migration has already started. Saturation's traction is the proof that producers will switch. The window is the migration, and it is open now.

Source: AICP. Labour share measured from a real AICP budget in our own sample set ($140,000 grand total; A+B+G+O = $56,044 = 40.0%).

The simple math, checkable in your head. One producer wins ten jobs a year at ~$150K = $1.5M of production. Crew wages at 40% = $600K. At a 1.49% payroll handling rate that producer's job flow is worth ~$8,940 a year in payroll fees — against $1,000 a year for the software seat. The attach is worth roughly nine times the subscription.

Our own network, fully converted: 100 producers × 10 jobs × $150K = $150M of production handled → payroll $894K + insurance $120K + seats $100K ≈ $1.1M a year from the personal network alone.

How big — sized bottoms-up, two markets not one.

Why ~3,500 is defensible. AICP represents 400+ production and post companies responsible for roughly 80–85% of nationally aired US motion-image advertising, in a $5B+ industry. An AICP industry survey puts responding companies at an average of 33 commercial productions a year. But AICP captures only the professional core: the Census counts 17,366 employer establishments in Motion Picture and Video Production, a category that explicitly includes television commercials — thousands of smaller shops doing regional advertising, social campaigns, DTC content and agency-direct work sit outside the AICP ecosystem entirely.

LPs travel, so these overlap. The geography is built independently of the segment estimate and lands in the same range, which is the check.

Comparable businesses — the pattern we are copying. The right comps are not budgeting tools. They are vertical operating systems for fragmented, unglamorous industries, and they share three traits: they began inside one workflow, they were founded by someone who had done the work, and their revenue eventually came as much from moving money as from selling software.

The product is live and real. 167 commits between 6 May and 12 July 2026. 310 unit tests. 25 database migrations, all in production. Next.js, Supabase, Anthropic. In beta at 7eight9.com.

What is actually shipped: the AICP 2019 builder with inline grid editing and drag-reorder; AI Quick Start and budget-aware chat with propose-and-undo; import from Showbiz CSV, Saturation JSON and legacy .xls; PDF, XLSX and CSV export in two layouts; version snapshots and comparison; client share links; collaborators with row-level security; and the full production lifecycle — estimate → award → running → actual → reconcile → sign-off, with change orders and a job-total rollup.

What the machine does: drafts the budget from a conversation, populates standard line items for the production type, applies fringes and markup, reconciles the topsheet, and flags what is missing. What stays human, permanently: what the job actually needs, the client conversation, and the decision to send it.

End of 2027 — ~$2M EBITDA, five people. 500 customers, including five production companies at ~$50M revenue each. Charging real money, priced on value: the land-grab phase ends and monetisation begins. Headcount: five at most — a CTO, and the rest customer service and finance/compliance. Best in class at making small companies capable of servicing big clients the way big companies do — estimates near-immediate, overages near-immediate, producers get their time back.

End of 2031 — a production machine. Not a budgeting company and not a payroll company: a kind of entity not yet seen, where AI does the bulk of the administrative and management work end to end. A job today needs an EP, a line producer, a production manager, a coordinator and often an office PA. Easy Budget turns that team into two people. Budgets self-actualise — data captured as the job moves from estimate to running to actual, pulled directly from card transactions and files rather than re-keyed. The books self-audit. Budgets self-populate payroll and fund the payout when the money is there.

Why the industry will let this happen. Agencies are dying and production companies are changing shape. Directors increasingly want to stay freelance. AI is making more possible. If we build tools working professionals use at scale and speed, we capture a meaningful piece of the industry as it reorganises — and brands will likely use the tool in-house directly, replacing much of what producers and production companies do today.

The 2031 number depends on how we price, and I am not going to pretend I know which it is. Three shapes, all from the same product:

Why the third one is the real answer. We are replacing skilled human labour at scale. An agency producer costs around $150,000 loaded. If Easy Budget lets an agency run with two producers doing the work that used to take five, that is roughly $450,000 a year saved. Charging $100,000 — the price of one of the heads saved — leaves the customer $350,000 better off and is an easy conversation.

Nine external producers surveyed (eight fully independent), across two waves — May 2026 pre-product, and August 2026. GX asks for about twenty; this is roughly half the bar.

They name the incumbent's failure themselves:

“It crashes sometimes. Sometimes it's tricky when 2 people look at it at the same time.”

Lars · Producer, Common Deer · 28 Aug 2026

“Visual Basic Errors. Being locked into an excel sandbox. Lack of easy visual side panel displaying all saved versions of a budget in chronological order.”

Connor MacDonald · Producer, Imposter · 28 Aug 2026

Connor bids more than 10 budgets a month.

“Feels dated and clunky at times. Quick revisions, alternate scenarios and comparing versions could be easier, and collaboration isn't as seamless as it should be. There's also a pretty steep learning curve for newer users.”

Adrian Rozas Cristino · Director, PIFF · 29 Aug 2026

“Have to pay for hotbudget every year, not easy to share with those not in the US… Should be a simple excel that anyone can use.”

Sarah Park · Producer · 5 May 2026

The most important finding in the dataset: producers do not ask for AI, they ask for speed. "AI-powered line-item suggestions" ranks last of ten features — 3 of 9. Yet two respondents described exactly what our agent does, without ever using the word:

“Ideally, I'd be able to enter the basic parameters of a project and have the tool generate a strong preliminary budget, including the standard line items for that production type. It could flag expenses I may have overlooked… When rebuilding a budget from scratch, the biggest risk is often not knowing what I've forgotten to account for.”

Matt Benson · Executive Producer · 28 Aug 2026

Matt bids more than 10 budgets a month, and offered to get on a call.

“Smart learnings from past shoots would be great. i.e. last time we budgeted this, we were under/over on these items.”

Alec Carrier · Executive Producer, Ride Also · 28 Aug 2026

So we sell the outcome, not the mechanism. AI-first is our architectural argument against Saturation. It is not the customer promise.

Nine of the ten most-wanted features are already live — version history (8/9), budget templates (8/9), PDF/Excel export (7/9), cost tracking vs actuals (6/9), real-time collaboration (6/9), cloud sync (6/9), client-facing share (5/9), mobile (4/9), AI suggestions (3/9). The one exception is integrated scheduling (4/9), which we have not built.

Pricing evidence. Six of nine would pay under $15/month; two at $15–30; one at $30–50. Nobody is anywhere near the $83/month that a $1,000-a-year seat implies. That is not a problem with the plan — it is the plan. See the economics.

Price today is a deliberate land-grab, not a valuation. Solo $10/mo, Studio $15/mo, Agency $79/mo — built in Stripe, never charged. An early user takes real risk on an unproven tool during a live bid; underpricing makes that risk feel like value, and the power gets discovered once they are inside. Narrow margin now is the point.

Where the money actually is. Producers told us they will pay under $15 a month. So the software stays cheap forever, on purpose — and the business is the attach. One producer's ten won jobs a year at ~$150K is $1.5M of production and $600K of crew wages; at 1.49% that is ~$8,940 a year in payroll fees versus $1,000 for the seat.

Cost to serve, measured. The product runs on `claude-sonnet-4-6` at $3 per million input tokens and $15 per million output. Peak usage — June and July 2026 — was under 300,000 tokens a month across the whole beta.

Where the revenue goes — and why 2027 has two possible shapes. Software runs a 97% gross margin, but almost all of the 2027 revenue is payroll, and payroll is an operating business rather than a software line. It carries banking and processing costs, multi-state tax filing, workers'-comp administration, and support headcount. So the revenue number depends entirely on whether we rent that operation or own it.

Payroll is not upside — it is the plan. Without it the two-year target is not reachable. With it, it clears. And the largest dependency in the model is already de-risked: our advisor Kenny Kline built a payroll company and has offered to help.

01 · Seats, from the beachhead into the firm. A freelance line producer proves it on a job; the production company standardises; team seats follow. The founder-level relationships at Softspot, Smuggler, Directors Bureau and Cabin Edit are firm-level entry points, not individual ones.

02 · The job lifecycle itself. A budget is not a one-time document — estimate, award, running, actual, reconcile, overages. Revenue attaches to the whole life of a job, and the product already ships all of it.

03 · The services funnel — the big one. Easy Budget sits at the exact moment a producer decides what a job costs. That is upstream of production insurance and payroll: large, recurring, high-margin businesses sold today by people who do not know a job exists until after it has been budgeted. Owning the budget means owning the referral. Wrapbook built a ~$1B valuation on production payroll; if we capture payroll for the jobs run through the platform, we can give the budgeting away.

The wedge is the bid. The commercial AICP budget, at a price that undercuts the market, for the freelance line producer who owns their own tool choice. Time to value is the same day: a producer talks to the agent and has a draft budget in the first session.

Why this wedge. It is the highest-frequency, highest-pain artefact in a producer's month — four bids a month, five hours each, most for jobs they will not win. It is also the document that everything downstream depends on: the award, the running budget, the actuals, the payroll and the insurance all descend from it. Own the bid and you are upstream of everything.

How the model resolves. Seats never become the business — producers have told us they will not pay software prices. Revenue resolves toward payroll and insurance attach, where one producer is worth ~9× their subscription, and where the customer relationship is already ours because we hold the budget.

Closed doors — what we said no to. Indie and documentary film, deliberately: different workflow, different buyer, no AICP. Movie Magic import, struck. And we are not selling software to bookkeepers-of-production — we own the customer relationship rather than licensing the engine to competitors.

The ramp is a sequence of gates, not a hockey stick. Each gate has a tripwire — a condition that must be true before we spend money on the next one. I am writing it this way because my failure mode is doing everything at once, and because these are exactly the decisions I want Gateway X's judgment on.

Where EBITDA actually comes from — the break-even is the whole argument. A fully-converted producer is worth ~$8,940 a year in payroll fees (10 jobs × $150K × 40% labour × 1.49%) against $1,000 for a seat.

Stickiness — the comps say operational, and the product is built for it. ServiceTitan and Toast are not defended by price; they are defended by the fact that once a system runs the money, leaving means retraining everyone and migrating years of history. Easy Budget is built the same way on purpose: a budget is not a document here, it is the job's system of record. Once a job is awarded, the estimate is frozen as a contract baseline and the running ledger, change orders and actuals all descend from it. You cannot move a live job to another tool mid-flight. Add payroll and the job's money runs through us too.

Twenty years inside the problem. I have produced broadcast commercials, socials and feature films, and been a partner at multiple regional and national commercial production companies. I have built these budgets myself, lost weekends to them, and watched what happens when a number is wrong in front of a client. Easy Budget is the tool I wanted and could not buy.

I am still the customer. My own company's real jobs run through it — six awarded budgets, eleven change orders — because I am willing to carry the accuracy risk myself before I ask anyone else to.

And I have built the apparatus before. Five years ago I moved to Nashville to help an old friend whose production company was contracting. His clients were local and regional. I outlined a plan to go after national brand work, and we won DoorDash as a heavy underdog. Then OpenDoor. That opened an agency relationship that awarded us a national Lululemon campaign — the biggest budget the company had ever seen. I loved closing it and I loved building the machine that delivered it. That is the same job as this one, at a different scale.

Followership. I am not a social media influencer or a thought leader. I am a reasonably smart person with good values trying to solve a set of problems that make me and some other people rich. But I hold a lot of trust and I have held it a long time: the directors on my roster trust me to represent them well and grow their careers; the producers and crews who work for me trust me to take care of them; the agencies and brands who hire my company trust me to deliver on budget, on schedule, and to give them something to be proud of. I can get a lot of people on the phone, and most people who work with me believe in my potential.

The evidence in this project specifically: Cody gave the beta feedback that became four shipped PRs. Kenny Kline advises for free and made the Gateway introduction. Maddie Clarke (founder, YadaYada) and Micah Ross (founder, Arden Productions) took beta invitations. And when I put a survey into my network in August, six named EPs and producers replied inside 48 hours — one of them with a 300-word product brief and an unprompted offer to get on a call.

When I was crew, my peers used to nominate me to go negotiate with the producers. That is the followership I would point at. Nobody assigned it.

Iteration receipts. Cody's four pieces of feedback each became a shipped PR — autocomplete of common AICP positions, one-click crew copy from prep, adding line items in running and actual, and a dashboard sort that folds shared budgets into one list. I scrapped and rebuilt the entire fringe re-deal model after reviewing my own first version as clunky. And when a producer thought post-production had been dropped from the maths, I audited the whole engine, proved the totals had always been correct, and fixed the cover that was showing fringe twice.

A decision I reversed on evidence. I set out to sell AI. My own demand data says producers rank AI-powered suggestions last of ten features — while describing, in their own words and unprompted, exactly what the agent does. So AI became the architecture argument and speed became the promise. People don't ask for AI. They ask for time.

What has to be true for me to still want to run this in five years: it has to be growing — and I have to still be obsessed with the product. Not obsessed with adding functions. Obsessed with it functioning at a higher and higher level. I want the people who use it to love it, and to want to show it to colleagues and friends, because it genuinely improves their lives and makes their work better and easier. If it becomes software sold to people who don't do the work, I will not want to run it.

What I want from Gateway X is not capital. This part of the journey is not particularly expensive — the cost to build the MVP can be contained, and it has been. Investment would let me hire out legal and technical work, and that is real. But it is not what I am short of.

I need structure. I have a hardcore work ethic, a network and an idea. What I do not have is a way of organising that effort into strategic tranches, and knowing how to gate what happens when — the tripwires that say this is the moment to open the beta, to move price, to hire, to turn on payroll. Left alone I will keep building, because building is the thing I am good at. That is exactly how I ended up with a product I trust more than I have customers who use it.

The thesis it produced: users + data = profit. Getting the professionals who oversee the billions of dollars moving through the advertising production economy onto one platform is the valuable thing. Monetising the construction of that platform through the tools they need today is what funds it — and being the platform is what lets us anticipate and build the tools they will need in an unknown tomorrow.

Adam · product. What to build first and what to refuse. I have shipped nine of the ten features producers asked for and have one repeat user outside my own company — building was never the constraint. The thing standing between me and an open beta is a Hot Budget importer with a provable ±$0.01 accuracy contract; I have specified it and not started it. I want your judgment on whether that is genuinely the unlock, or whether I am engineering my way around an activation problem. And I would like to be talked out of integrated scheduling early rather than halfway in.

Nak · the operating machine. The tranches and the tripwires — that is the ask, and it is the main one. Plus: instrument the funnel properly (I had to write SQL to learn that 21 signed up, 8 built something and 6 jobs were awarded, and I have no cadence for watching it); apply your standard on scorecards and reporting rhythm; and help me design, source and vet the ops-and-finance hire, which is the honest archetype gap in my team. Then the largest strategic decision in the business: payroll, build or partner. Kenny will help on domain. I need help on the decision.

Jesse · GTM and leadership. We have not met yet. When we do: my distribution is a real unfair advantage with a ceiling I can see — 100+ producers, founder-level relationships at four companies, and the producers I hire today for brands including Coca-Cola and Adobe. That reaches the first fifty customers. It does not obviously reach the next five hundred. I need the second channel designed, and I need the monetisation sequenced: producers say they will pay under $15 a month, my model needs $1,000 a seat a year, and the resolution is the attach. What I lack is the trigger and the date.

What I bring. Nobody is positioned as well as I am to build in this space. It is a niche industry with a great deal of money moving through it, and a genuine entry point to capture multiple revenue streams by building an integrated suite of tools that help professionals do their job better, faster and easier. I know the things I know and can achieve, and I know the things I cannot.

I am more interested in mentorship, accountability and thought partners than in capital. I like Adam and Nak. I trust Kenny. That is why I am here rather than raising.