An on-sale is not a launch. It is a measurement. Within seventy-two hours a show has told you, in ticket-level detail, how much latent demand existed, how concentrated it is, which price bands carry it, and how much of the run will need to be built rather than harvested. Almost every expensive mistake in live marketing comes from treating those three days as a revenue event instead of a data event.

Three shapes, three different runs

Plot cumulative sales against time and most on-sales fall into one of three shapes.

The spike. A steep first hours, then a near-flat line. Existing demand was large and already addressable: the fanbase knew, the announcement worked, the list was warm. The spike is good news about awareness and bad news about the remaining run, because the easy tickets are gone and everything after this point costs more per seat. The mistake here is reading the first-day efficiency as a baseline and budgeting against it.

The ramp. A modest start that builds over days. This is a show being discovered rather than awaited. Ramps reward patience and punish early panic. Cutting spend on day two because the first-day number looked soft is the most common way to turn a healthy ramp into a flat line.

The stall. A reasonable opening that stops. Something is wrong upstream of the campaign: the date, the venue, the price, the market, or a competing event nobody mapped. No amount of budget fixes a stall, and pushing it is how a marketing problem becomes a pricing problem.

What to measure, and at what grain

Aggregate revenue over the first three days is almost useless. The decisions that matter need a finer grain.

• Sales per hour, per price band. A top-band sell-through with a soft mid-band means the pricing ladder is wrong, not the targeting.
• Percentage of capacity, per date. In a multi-date run the aggregate hides everything. One date at sixty percent and one at fifteen is a completely different problem from both at forty.
• Source mix against sold tickets, not reported purchases. The platform ledger will overcount. The box office will not.
• Geography against travel radius. Sales concentrated outside the usual catchment usually mean the local market has not been reached yet.

The decisions that belong in week one

Three, and only three. First, whether the run needs building or harvesting, which determines whether budget grows or holds. Second, which dates are structurally weak and need their own treatment rather than a share of a run-level budget. Third, whether the price ladder is doing what it was designed to do, because a ladder that is wrong on day three is still wrong in week six, only with less inventory left to fix it.

Everything else can wait. The curve will keep talking.

Why this is a data problem, not a reporting problem

Reading the curve properly requires ticket-level sales joined to spend at the same grain and on the same clock. Most organisations have both numbers and no join: sales in the ticketing system, spend in three ad platforms, and a spreadsheet reconciling them on Monday. By Monday the first seventy-two hours are gone and the decisions were made on instinct.

The NYBA platform reads sales from the ticketing systems already in use and puts them next to spend in real time, per show, per band, per date. The point is not a nicer dashboard. It is that the on-sale curve becomes readable while it is still actionable.