What OTS means, and why most vehicle advertising numbers are wrong

In August 2026 we rebuilt the engine that estimates how many people see a Velo4U campaign. The headline number it produces fell by roughly a factor of nine. We are publishing the whole rebuild, including the part where our old numbers were wrong, because the alternative, quietly changing them, is what the rest of this industry does and it is exactly why nobody believes any of these figures.

The measurement chain, as it is actually defined

The Out of Home Advertising Association of America publishes the OOH Impression Measurement Guidelines, the closest thing this medium has to a standard. It defines three distinct quantities, and the distinction between them is where nearly all inflation happens:

OTS is the trading currency. When a media buyer asks for impressions, this is the number they mean. The common failure is to compute circulation, label it impressions, and never mention the viewshed step at all, which is precisely what our own engine did until we caught it.

What our old numbers were doing wrong

The previous engine used 250 vehicles per kilometre on freeways, 120 on arterials and 50 on local roads, an occupancy of 1.38 people per vehicle, and no viewshed step whatsoever. Those constants were not sourced from anything. They were plausible-looking numbers, and they produced about 150 OTS per kilometre.

Three separate problems, compounding:

1. The road volumes were three to twenty-five times too high. Measured against real traffic counts, no Melbourne road band came close to those figures.
2. The occupancy figure was above the all-purpose average and was being applied to commute-heavy driving, which runs well below it.
3. The output was mislabelled. With no viewshed factor, what the engine produced was circulation. We were calling it OTS.

Rebuilding it on Victorian traffic data

Step one: real road volumes

We took the Victorian Department of Transport and Planning’s Historical Annual Average Daily Traffic Volume release for the declared road network, 9,959 Melbourne-area segments covering 12,381 kilometres. Oncoming vehicles met per kilometre is 2 × (q ÷ v), where q is daytime hourly flow and v the operating speed. Computed straight off that dataset:

Local streets sit outside the declared network entirely. At roughly 700 AADT per direction they contribute about 2 vehicles per kilometre, and VicRoads puts around 75% of metropolitan vehicle-kilometres on arterials and freeways, so local roads carry the remaining quarter.

On top of oncoming traffic we add 15% for same-direction overtaking. Only 15%, because traffic travelling with you is one long exposure, not a fresh one every kilometre.

Step two: people, not vehicles

Melbourne’s journey-to-work occupancy is 1.09 people per vehicle (VISTA 2015–16). All-purpose travel runs higher because non-commute trips carry passengers. We use 1.2, the mixed-purpose figure. The old 1.38 was above even the all-purpose average.

Step three: the viewshed discount we had been skipping

This is the step that separates circulation from OTS, and the one that matters most for vehicles specifically. A door panel is roughly half a square metre. A billboard is 12 to 45 square metres. Most encounters with a campaign car are oncoming vehicles passing at a combined 90 to 110 km/h, at an oblique angle, frequently screened by other traffic.

So the viewshed factor varies by road type, and it is not generous:

• Freeway, 0.30. High speed, wide separation, often a median in between.
• Arterial, 0.40. The blend case: two in five people who pass the car were positioned to see the ad.
• Local, 0.55. Slow, close, frequently stationary.

What the rebuilt engine actually produces

Putting the three steps together, per kilometre driven, before any time-of-day adjustment:

The blend assumes a realistic distance mix of 20% freeway, 55% arterial and 25% local. It lands at roughly 47 people per kilometre in circulation and 17 in OTS, a 64% discount at the viewshed step, and around nine times below the 150 per kilometre we used to quote.

Every Velo4U coverage report now prints both numbers. Circulation and OTS, side by side, so the discount is visible rather than buried. Parked exposure is reported on its own line and never folded into the headline.

Why we published a number that makes us look worse

It does make us look worse on one axis. A campaign that used to claim hundreds of thousands of impressions now claims tens of thousands, and on a naive cost-per-thousand comparison that repositions vehicle advertising from a cheap impression to a premium one.

We think that is the correct position, for two reasons.

The first is that the impressions are qualitatively different. A car on a Werribee school run is seen by the same local audience repeatedly, at low speed, in the suburb where they actually spend money. A freeway billboard impression is one glance from someone driving out of the area at 100 km/h. Pricing those identically per thousand was never right in either direction.

The second is simpler. If we quote you a number and it is inflated, you find out, not immediately, but eventually, when the campaign does not produce what the arithmetic promised. There is no version of this business that survives that. We would rather sell a smaller number that holds up.

Four questions for any outdoor vendor

1. Is this circulation or OTS? If they do not know the difference, it is circulation.
2. What viewshed factor did you apply? There should be a number, and it should be well below 1.0 for anything smaller than a billboard.
3. Which traffic dataset are the road volumes from? It should have a name and a year.
4. Is the distance measured or assumed?“Drivers average 250 km a week” is an assumption. A GPS log is not.

Ours are: OTS, published alongside circulation; 0.30 / 0.40 / 0.55 by road band; Victorian DTP AADT, 2019 release; and every kilometre GPS-measured, never assumed. The full derivation lives in the codebase that generates your report, and the constants in this article are the constants that engine runs on.

Common questions

Related: what car advertising costs in Melbourne · car advertising vs billboards · how a Velo4U campaign works