This page defines the terms used throughout the Mira attribution documentation, along with the industry vocabulary you are likely to encounter in an out of home (OOH) measurement conversation.

Terms are grouped by the part of the process they belong to rather than listed alphabetically, so that reading a group start to finish explains a stage of the process. Use your browser's find function to jump to a specific term.

📘
Start With the Overview
If you are new to OOH attribution, the Overview covers the same ground as a narrative rather than a reference.

Study Anatomy

The objects and parameters that define what is being measured.

Term	Definition
Advertiser	An organizational container in the Mira Console, used to group the resources belonging to one brand. Deleting an Advertiser does not delete the resources associated with it. See Advertisers.
Attribution	The process of determining whether media exposure had an effect on a conversion or other KPI.
Attribution Window	The amount of time a user has to convert after being exposed in order to be counted as an attributed conversion. Also called a lookback window or conversion window. A user exposed at `2020-12-01 10:21:00` under a 7-day window is a converted user only if a conversion event falls between that moment and `2020-12-08 10:20:59`.
Campaign	The media buy being measured. A Campaign holds the dates, the associated Sources, and one or more Studies. See Campaigns.
Cut	A subgroup of a study's results - the practice of grouping results into smaller sets and observing the effect within each. A cut key may be a standard metadata field such as `creative_id`, a custom metadata field, or a label applied when a Source was associated with the Campaign. Market, property, and goal cuts are always included; any requested cut is in addition to those.
Exposure-Side Cut	A cut on a factor of the media, such as market, vendor, format, or creative. Exposure-side cuts reduce both the numerator and the denominator of a conversion rate, because they narrow the set of people under consideration.
Conversion-Side Cut	A cut on a factor of the conversion, such as property or goal. Conversion-side cuts reduce only the numerator, because they narrow only the set of qualifying conversion events.
Flight	A distinct run of media within a Campaign. A Campaign's date range must cover all of its flights.
Incremental Lift Analysis	Mira's standard measurement product, and the only Study type. See Incremental Lift. It compares the change in conversion rate from the pre period to the post period in the exposed group against the same change in the control group. The pre period and the control group are each optional configuration on the Study.
Post Period	The period from the start of the media campaign to the end of the media campaign plus the attribution window. Exposures occur only during the campaign itself; the trailing stretch exists to let the attribution window lapse for users exposed near the end.
Pre Period	A period before the start of media, used to measure a baseline conversion rate for each group. The pre period lasts as long as the attribution window. A Study is only eligible for a pre period if its conversion environment was collecting data before the pre period began.
Study	The measurement configured against a Campaign: its type, dates, attribution window, cuts, and whether a pre period and a control group are requested. One Campaign may have more than one Study - for example, the same campaign measured at two different attribution windows.
Study Status	Where a Study sits in its processing lifecycle: `PENDING` before it begins, `IN_PROGRESS` while results are being computed and refreshed, `DELIVERED` once it has stopped updating, and `ERROR` if it could not be completed.
Approval Status	Whether a Study has been submitted for measurement and accepted: `UNSUBMITTED`, `SUBMITTED`, `APPROVED`, or `REJECTED`. Distinct from Study Status, which tracks processing rather than approval.

Exposure

Determining which users saw the media. See Sources and Exposure.

Term	Definition
Ad Roll Log	A record of the exact times a creative played on a digital OOH unit. Required for digital OOH measurement, and not the same thing as a proof of play. See Digital.
Digital OOH	OOH inventory that rotates between multiple creatives, so that the media being measured is not always on screen. Measurement requires ad roll logs to establish when the campaign's creative was playing.
Exposure	An observed opportunity to see the media: a device recorded inside a unit's viewshed during a qualifying time window.
Exposure Zone	An 800 m circle drawn around an OOH asset, used in place of a viewshed where a jurisdiction restricts the sale of precise location data. Larger and non-directional, so the exposed group it produces contains false exposures and results are attenuated. Opt-in per Study. See Exposure Zones for Coarse Regimes.
Exposure File	The output of the exposure process - a record per exposure, carrying an advertising ID, a timestamp, and any metadata available for exposure-side cuts.
Facing	The direction an OOH unit is oriented, given as a cardinal or intercardinal direction such as `N` or `NW`. Used to shape the viewshed.
Format	The type of OOH inventory, such as a bulletin, a poster, or a transit shelter. Affects the size and shape of the viewshed and is a common cut.
Geofence	A two-dimensional shape on the map, expressed as a polygon or multipolygon. The general term; a viewshed is the specific geofence around an OOH unit.
Mobility OOH	OOH inventory that moves, such as a vehicle wrap or a transit-mounted screen. Position always varies; content may be fixed or rotating. Measurement requires high-resolution waypoint logs carrying position and time, plus creative play windows where content rotates. Also called moving OOH. See Moving.
Proof of Play (POP)	A vendor report demonstrating that an order was fulfilled, typically aggregated to counts per unit per hour. Because it does not record the exact time of each play, a POP cannot be used in place of an ad roll log.
Source	A container for OOH source data, with a type - `STATIC`, `DIGITAL`, `MOVING`, or `STATIC_GEOJSON` - and a configuration describing how to interpret files placed in it. Sources are associated with Campaigns to tell Mira where a campaign ran. See Sources.
Source Job	A single unit of ingest work: one file or one scheduled delivery arriving into a Source.
Static OOH	OOH inventory that displays the same media for the duration of the campaign, such as a roadside billboard. When measuring static sources, the time component is ignored apart from the campaign or flight dates. See Static.
Unit	A single piece of OOH inventory, identified by `unit_id` in a source file.
Vendor	The publisher or media owner operating the inventory. A common cut, and the party who supplies ad roll logs for digital inventory.
Viewshed	The area around an OOH unit within which a person is considered to have had the opportunity to see it. Derived from the unit's coordinates, format, size, and facing. Contrast with an Exposure Zone, the much larger stand-in used where precise location data is unavailable - see Exposure Zones for Coarse Regimes.
Waypoint	A timestamped position report from a moving OOH asset, carrying a latitude and longitude and usually a start and end position for each interval.

Groups and Control

The two populations a study compares.

Term	Definition
Control Group	The set of users who were not exposed to the campaign, used as the counterfactual. Because OOH exposure cannot be assigned in advance, the control is constructed after the fact to resemble the exposed group across demographics, fidelity, and geography. See Control.
Exposed Group	The set of users observed inside a unit's viewshed during the campaign. Also called the test group.
Fidelity	How much distinct information is observed for a single device - the depth of the location data, as opposed to its breadth. Fidelity is one of the dimensions on which the control group is matched to the exposed group.
Matching	Constructing the control group so that its distribution across the matching dimensions resembles the exposed group's, so that the two differ in exposure and as little else as possible.
Sample	The portion of the true exposed population that Mira observes. OOH attribution always works from a sample, which is why hypothesis testing is fundamental to it. See OOH Differences from Digital.

Conversion

The events being attributed to exposure. See Conversion Properties.

Term	Definition
Brand Survey	A survey fielded to exposed and control users to measure attitudinal outcomes such as awareness, ad recall, favorability, consideration, and intent. Because responses are actively solicited during the campaign, brand surveys must be configured before it begins. See Brand Lift Surveys.
Conversion	A qualifying event by a user who was exposed, occurring within the attribution window.
Conversion Event	A single row of conversion data: a user identifier, a timestamp, and metadata describing what happened.
Conversion Property	A scoped source of conversion event data, of type `WEB`, `SDK_POSTBACK`, or `FOOTFALL`. A property is associated with a Study to tell Mira what to measure conversion to.
Footfall	Using Mira's location data itself as the conversion environment, so that a visit to a physical location counts as a conversion. Also called walk-in conversion. Because the underlying data is always on, footfall studies can be run retroactively. See Footfall Locations.
Goal	A named conversion action within a property - a pixel on a web property, or an SDK event on an app property. Goals are the unit at which conversion is reported and cut.
Goal Health	Whether a goal is currently receiving event data, reported in the Console as healthy, warning, or unhealthy. See Understanding Goal Health.
MMP (Mobile Measurement Partner)	A third-party analytics platform embedded in a mobile app, such as AppsFlyer, Kochava, or TUNE. App conversion measurement works by configuring the MMP to post events back to Mira's data provider.
Pixel	A small, invisible image embedded in a web page. Loading the page requests the image, and the request is logged with the goal, the property, and the user's identifiers - which is the basis of web conversion data. See Web Pixels.
SDK Postback	An app event forwarded from an MMP to Mira's data provider. All events should be configured to post back, not only those the MMP itself attributed. See App SDK Postbacks.
Whitelist	The set of goals declared on a property - the pixels or SDK events that will be tracked and reported for it.

Metrics and Statistics

How a result is expressed and how much confidence to place in it.

Term	Definition
Benchmark	A reference value from a comparison set of past studies, used to contextualize a result. Benchmarks are keyed by vertical and property type.
Confidence Interval	A range around a reported metric that quantifies its uncertainty. Mira reports a 95% confidence interval around net lift.
Confidence Level	How confident we are that an observed result reflects the whole population rather than the particulars of the sample. A confidence level must be reported alongside a result for that result to be usable.
Conversion Rate	The number of unique users who converted, divided by the number of unique users observed.
Hypothesis Testing	The statistical procedure that determines whether an observed difference between two groups is large enough to be distinguished from sampling noise.
Lift	The relative change between two conversion rates. A rate moving from 1% to 2% is a lift of 100%.
Net Lift	The difference between the lift the exposed group experienced from pre to post and the lift the control group experienced over the same span. The primary result of an Incremental Lift Analysis. See Incremental Lift.
Statistical Significance	Whether an observed result is unlikely enough to have arisen by chance that it can be treated as real. Depends on the sample size, the number of conversions, and the size of the underlying effect - which is why it can never be guaranteed in advance. See Study Feasibility.
Vertical	The industry category of a Campaign, such as retail or automotive. Determines which benchmark a result is compared against.

Identity and Data

The identifiers and datasets the process runs on.

Term	Definition
Advertising ID	The identifier a device presents to advertisers. See MAID.
Device Graph	A dataset linking identifiers that belong to the same person or household, such as advertising IDs, cookies, and IP addresses. Used to connect an OOH exposure to a conversion that happened on a different device or in a browser.
Household Counting	Counting unique households rather than unique devices, by resolving device identifiers to a household through the device graph. Changes the denominator of every conversion rate in a study, and excludes exposed devices that do not resolve.
Imprecise Location Data	Location data that has been coarsened, so that a reported position describes an area rather than a point, because the jurisdiction restricts the sale of precise geolocation. Exposure for it is computed against an Exposure Zone rather than a viewshed. See Exposure Zones for Coarse Regimes.
Location Event	A single observation from the mobile location dataset: an advertising ID, a timestamp, and a latitude/longitude pair.
MAID (Mobile Advertising ID)	The advertising identifier assigned to a mobile device - IDFA on iOS, GAID on Android. The join key between exposure and conversion data.

Geography

The geographic objects that appear in cuts, filters, and source data.

Term	Definition
Block Group	A US Census geographic unit, smaller than a postal code. Available as a cut and as a filter on POI-based segments.
CBSA	Core Based Statistical Area - a US Census metropolitan or micropolitan region.
Covering	A representation of a polygon as a set of discrete S2 cells at a chosen precision. Coverings are what geographic queries actually run against.
DMA	Designated Market Area - a Nielsen-defined media market. Mira's Markets correspond to DMAs, and market of exposure is a standard cut.
GeoJSON	An open standard for encoding geographic shapes. Used to supply exact viewshed or footfall polygons where a radius around a point is not precise enough.
Geopath	The OOH industry's measurement body, formerly the TAB. Publishes unit-level identifiers and impression estimates; `geopath_id` is a standard metadata field on OOH source data.
Market	Mira's representation of a Nielsen DMA, carrying both the Nielsen identifier and the Geopath/TAB market key.
POI (Point of Interest)	A location of interest such as a retail store or restaurant, held as a polygon with associated metadata.
POI Category	A class of POI, such as Veterinarian or Restaurant. A POI may belong to up to three categories.
POI Chain	A brand or franchise grouping POIs together, such as all locations of one restaurant chain. A POI belongs to at most one chain.
Postal Code	A US ZIP code, available as a geographic filter and cut.
S2 Cell	A unit in Google's S2 Geometry system, which maps the Earth's surface to a hierarchy of cells at 30 levels of precision. Gives every location an integer identifier and makes containment queries cheap.
WGS84	The coordinate system used throughout the platform. Positions are given as a `(longitude, latitude)` pair in decimal degrees.

Industry Measurement

Terms that come from OOH planning rather than from Mira, and that often appear alongside attribution results.

Term	Definition
Frequency	The average number of exposures per reached individual.
GRP (Gross Rating Points)	Reach percentage multiplied by frequency. One GRP is 1% of the target audience reached once.
Impressions	The estimated total number of exposure opportunities a buy delivers - reach multiplied by frequency. Impression estimates are the basis of Mira's feasibility guidelines.
LTS (Likelihood to See)	Opportunity to see, adjusted for visibility factors such as format, size, viewing angle, and approach speed.
OTS (Opportunity to See)	The count of people whose path crossed an asset's viewshed, before any adjustment for whether they actually looked.
Reach	The number of unique individuals exposed at least once. Because Mira's analysis counts unique users, its figures are closer to a reach metric than to an impression count.
TRP (Target Rating Points)	Targeted impressions as a percentage of the market population, including duplication.

📘
A Note on Unique Users
Mira's conversion rates are built on unique users, not impressions. This is the most common source of confusion when comparing a Mira result to a media plan: a campaign delivering tens of millions of impressions will report a far smaller number of observed users, because each user is counted once regardless of how many times they were exposed. See the FAQ for more on this.

Glossary

Start With the Overview

Study Anatomy

Exposure

Groups and Control

Conversion

Metrics and Statistics

Identity and Data

Geography

Industry Measurement

A Note on Unique Users

What’s Next