From one lane to the whole venue: the OS roadmap.

Strip any venue back to its bones and you find the same five lanes of work: sell the space, run the bookings, count the money, manage the stock, staff the shifts. Five lanes — and in most independent venues, one owner or GM stretched across all of them, plus whichever admin hires the margins can bear.

We call MarsAI a venue operating system, and that phrase is a commitment, not a metaphor. An operating system is not a tool you use; it's the thing your venue runs on. Today that claim is true for exactly one lane — bookings and the enquiry admin behind them. This post is the roadmap for the rest: which lane comes next, what each phase actually replaces, and how we label the difference between shipping today and on the roadmap. We're precise about that difference, because a roadmap sold as a feature list is how this industry earned its scepticism.

The rule that decides what gets automated

Every phase on this roadmap is admitted by the same test. It's the test we already apply inside the booking lane, and it doesn't bend for ambition:

A lane is ready for the OS when three things are true: the work is process, the rules can be written down, and the data the lane needs already lives in the system. Everything that fails the test stays human — permanently, not temporarily.

That third condition is why this roadmap has an order at all. You can't forecast stock without knowing what events are confirmed. You can't build a rota without the events calendar. You can't run re-engagement marketing without knowing which enquiries went cold and why. Every later lane reads from the booking record — which is why bookings had to come first, and why point tools that start anywhere else stall.

Phase 1 — Bookings and enquiry admin Shipping today

The lane where revenue enters the building, and the lane most obviously shaped like a process: read, sort, reply, chase, log, hand off, report. In most venues it's a £30k+ role — a booking manager or events coordinator, plus the back-office admin behind them — or it's you, at midnight.

What it replaces: the booking-manager and enquiry-admin role — every channel read, every enquiry triaged by value, answered in your voice, chased until it resolves, logged and reported. What you keep: quotes above your thresholds, discounts, exceptions and sensitive replies all stop and wait for you.

This phase is live now as a managed deployment — from £995 a month, most venues £1,250–£1,750, against a fully-loaded hire cost of £2,900+ a month. The engagement starts with a free audit and a free two-week pilot on your real booking lane; the numbers are on the pricing page.

Phase 2 — Marketing On the roadmap

Marketing is next because the booking lane generates its fuel. After months of running your enquiries, the system knows things no agency deck does: which channels produce bookings rather than noise, what your seasonal demand curve actually looks like, and — most valuably — every enquiry that went cold and the exact point it stalled.

What it replaces: the marketing admin loop. Re-engagement campaigns to lost enquiries built from real pipeline history, seasonal pushes timed to your booking curve rather than the calendar's, channel budgets steered by conversion data instead of instinct, and drafts written in the same house voice the booking lane already uses. For many venues that's a chunk of an agency retainer and a shelf of email tools doing less work than they bill for.

What you keep: the brand. Nothing goes out — no campaign, no offer, no spend — without your approval. The OS does the assembling and the timing; you stay the editor-in-chief.

Phase 3 — Finance and admin On the roadmap

Every confirmed booking drags a paper trail behind it: deposit requests, balance chases, invoices, reconciliation, and the month-end scramble to assemble numbers for your accountant. It's the least loved admin in the building, and it's process from end to end.

What it replaces: the bookkeeping hours between your venue and your accountant. Deposits requested and chased on schedule, invoices raised from the booking record (the numbers are already in there — nobody should be retyping them), per-event revenue tracked against what was quoted, and a clean month-end pack handed to your accountant instead of a shoebox of PDFs.

What it doesn't replace: your accountant. Tax, filings and financial judgement stay with a professional. The OS replaces the admin either side of them. What you keep: refunds, write-offs and anything over your thresholds wait for your sign-off.

Phase 4 — Stock On the roadmap

Stock is where venues quietly bleed: over-ordering before a weekend that doesn't materialise, running dry mid-event, and the weekly count-and-order ritual that eats a manager's afternoon. What makes it automatable isn't clever forecasting — it's that by this phase, the OS already knows your calendar.

What it replaces: the count-and-order loop. Expected consumption tied to confirmed events and covers, par levels maintained per line, draft supplier orders raised ahead of demand, and wastage patterns flagged instead of buried. What you keep: orders wait for your approval — or auto-approve below a value you set, exactly like quote thresholds in the booking lane.

Phase 5 — Staffing On the roadmap

Last, deliberately. Staffing touches people's lives and pay, so it ships with the most conservative approval defaults of any phase. But the admin layer of staffing — the rota Tetris — is pure process, and by this phase the OS knows exactly what each week holds.

What it replaces: rota drafting from the confirmed events calendar, shift offers and confirmations chased by message the way follow-ups already are, gaps escalated early instead of at 4pm on Friday, and worked hours compiled for payroll. What it will never touch: hiring, firing, pay decisions, performance — anything about a person rather than a shift stays with you, forever. That's not a phase-one caution; it's the design.

Why this order, and why point tools can't get there

Read the phases again and you'll notice each one leans on the record built by the ones before it. Rotas need the events calendar. Stock needs covers. Marketing needs conversion history. Finance needs booking values. Bookings needed nothing but your channels — which is why it went first, and why it had to prove itself before anything else earned the right to ship.

This is also the structural reason a stack of point tools never becomes an operating system. Each tool starts from its own empty database and stays blind to the others, so a human ends up as the integration — retyping, reconciling, ferrying context between tabs. We wrote up what that costs in the software stack post, but the short version: every additional tool adds a lane to check. Every additional OS phase removes one.

What never gets automated

The through-line of every phase is the same executive list, and it only ever grows more explicit:

• Final say on quotes, discounts and exceptions above your thresholds
• Complaints, refunds and anything on your sensitive list
• Brand: voice, offers, and what your venue puts its name to
• People: hiring, pay, performance, and every conversation that matters

The OS runs the venue's process. You run the venue. If a vendor promises to automate the judgement too, they're describing a black box — and "fully autonomous" is the wrong goal.

Where this is honest

Today, the booking lane is the product. Phases 2–5 are the build path — developed with deployment venues, switched on one lane at a time, each piloted on a real operation before anyone pays for it. When we say roadmap, we mean roadmap: not a hidden feature flag, not a sales promise with an asterisk.

If you want the full-venue OS, the path there starts in the same place for everyone: the booking lane, because it's live, it's the lane that pays for the rest, and it builds the record every later phase reads from. The free audit prices the role you'd be replacing — and if the honest answer is "hire a human", that's what the written plan will say.