A growing number of US states restrict the sale of precise geolocation data. In those states Mira still receives location data, but coarsened - accurate enough to say roughly where a device was, not accurate enough to say which side of a street it was on.

Standard exposure does not survive that. A viewshed is a tight, directional shape, often a few hundred feet across, and matching a coarsened position against it produces noise rather than signal.

Exposure Zones are the response: instead of a precise viewshed, an 800 metre circle is drawn around each asset, and a device seen inside that zone is counted as exposed. It keeps campaigns measurable in these states. It also changes what an "exposed" device means, which is why it is opt-in and why results built on it must be read differently.

🚧

This Is a Deliberate Approximation

An Exposure Zone study is not a viewshed measurement with slightly worse inputs. It is a different, coarser method that trades precision for the ability to measure at all. Everything on this page about how to read the output follows from that trade, and none of it should be treated as a caveat to skim.

What the Regulations Restrict

These laws restrict the sale of precise geolocation data. They do not restrict its collection, and they do not restrict measurement as an activity - the constraint arrives upstream, in what data can lawfully be supplied to us.

"Precise" is defined by an accuracy threshold, generally 1,750 feet (roughly 533 metres). Data coarser than the threshold remains lawful to sell, which is what keeps any measurement possible. Thresholds are set state by state, so the specific figure can vary.

Where This Applies

Most states have adopted or are considering variations on a common model bill rather than writing their own frameworks, so new states tend to resemble the existing ones.

This table is the canonical list. Feasibility links here rather than repeating it, so only one page needs updating when a state changes status.

🚧

This List Changes

New Jersey passed with no advance notice and took effect immediately, which broke measurement across two major DMAs overnight. Treat this table as a snapshot and confirm current status with your account team before committing to a measurement plan - particularly for campaigns being sold months ahead.

How Exposure Zones Work

The change is confined to one step. Everything downstream - control construction, the attribution join, the lift calculation - is structurally unchanged.

A device seen inside the zone during a qualifying time window is counted as exposed.

The difference in scale is the whole story. An 800 m circle covers about 2 square kilometres - many city blocks, several streets in every direction, and a great deal of ground from which the asset is not visible at all. Many of the devices it captures were never in a position to see anything.

The zone is also not directional, where a viewshed is. A billboard facing north has a viewshed extending north; its Exposure Zone extends equally in every direction, including behind it.

Coverage Within a Campaign

The two regimes coexist. A campaign spanning several states will have precise units in unaffected states and coarse units in restricted ones, and exposure is computed appropriately for each. Every exposure carries a flag recording which regime produced it, which is what makes the choice below possible.

This matters more than it sounds, because DMAs do not respect state lines. A campaign bought in the New York or Philadelphia DMA has units on both sides of a ban boundary.

What It Costs

False Exposures and Attenuation

The zone captures devices that were near an asset but never able to see it. Those devices enter the exposed group as false exposures, and they behave like unexposed people, because that is what they are.

The consequence is attenuation: measured lift is pulled toward zero. A campaign with a genuine effect will read as a smaller effect, and often as no effect at all.

🚧

A Null Result Does Not Mean the Campaign Did Not Work

With Exposure Zones, a null is the expected outcome even for effective campaigns. It means the effect was not detectable at this data precision. Reading a null as evidence the campaign failed would be a serious misinterpretation, and it is the single most important thing to carry into any conversation about these results.

The size of the attenuation depends on how densely the area around each asset is trafficked, so it varies by market and by placement. It is best understood as substantial rather than marginal.

Confidence Intervals Are Wider

Misclassification adds uncertainty on top of ordinary sampling error, so intervals on an Exposure Zone study are meaningfully wider than the same campaign would produce on precise data. Treat the interval, not the point estimate, as the result.

Cuts Are Limited

Results should be read at campaign level only. Format, creative, and market breakdowns are not reliable here.

The reason is specific: the rate of false exposures is not uniform. A large zone around a highway bulletin and one around a bus shelter are wrong in different proportions, because the surrounding foot and vehicle traffic differ. A cut by format inherits that unevenness and would present it as a difference in performance.

Mixed-Regime Campaigns Blend Two Things

Where a campaign spans both regimes, an overall figure blends a well-measured effect with a heavily attenuated one, weighted by how the units happen to split. That blend is specific to the campaign, so figures from mixed campaigns are not cleanly comparable with each other or with precise-only studies.

There are two defensible readings, and the choice should be made deliberately:

• Exclude the coarse portion and measure the precise units cleanly, accepting that part of the buy is unmeasured.
• Include it and read the blended figure with the attenuation disclosed.

Other Effects Worth Knowing

Dense markets may not leave a usable control. Where a campaign has many units in a small area, the zones can merge to cover most of the urban core. The control group then draws from outside it, which makes the comparison partly about where people live rather than about exposure. This is assessed per campaign before a study is committed.

Concurrent campaigns overlap. Large zones from different campaigns overlap in ways tight viewsheds do not, so in markets with several buys running at once, attributing an exposure to one specific campaign is less clean.

Opting In

Because this is a different measurement, it is never applied silently.

Coarse exposures are excluded by default. Including them is a setting on the individual Study or export, so the decision is made per deliverable rather than per account:

• A Study can be configured to include coarse exposures, or to run precise-only.
• A data export can likewise include or exclude them.
• The same campaign can be read both ways, since the underlying exposures retain their regime flag either way.

📘

Which to Choose

If the campaign has meaningful spend in restricted states, including coarse exposures is usually the only way to measure that portion at all - accepting attenuation and reading the result as directional. If restricted states are a small share of the buy, a precise-only study will produce a cleaner, more comparable read of the rest. Your account team can advise on the split for a specific campaign.

Reporting on These Results

When results include coarse exposures, say so wherever they are presented. The output is an Exposure Zone study, not a viewshed measurement, and it should be described that way - with the attenuation, the campaign-level restriction, and the meaning of a null stated alongside the numbers.

See Feasibility for what is available where, and Incremental Lift Reports for reading lift figures generally.