The Measurement Ladder: Traffic, Viewability, and Audience
The subject here is the OOH measurement ladder: the sequence that turns a location count near a billboard into a more qualified statement about exposure and audience. The practical question is not which label sounds most impressive, but which conditions a reported number has actually met. Once the conditions are separated, readers can classify a metric before using it for planning, buying, or comparison.
From traffic to a functioning display
Start with the easiest observation. Location Traffic counts vehicles crossing defined roadway boundaries over a period. Location Circulation counts people, pedestrians, people in vehicles, or proxies such as mobile or transit activity in a defined location. The two may be useful for describing the scale of a site, but they are not identical to the population that can see a particular billboard. A roadway boundary can include vehicles traveling away from the display, people behind an obstruction, or people who pass through the area without entering the asset’s media-specific exposure zone.
The distinction is not merely technical vocabulary. A circulation figure is usually duplicative: the same person may appear in the count on multiple trips, and a single vehicle may represent more than one person. A location count may also come from several sources whose observations overlap. The MRC OOH standards therefore treat Presence as a condition that must be documented when circulation or traffic is used to qualify impressions or audience. They also call for attention to time of day, day of week, operating hours, seasonality, timing alignment, and the quality of third-party data. [MRC]
That makes traffic a useful starting point but a poor stopping point. It can describe potential scale, help compare sites, or serve as an input to a later model. It cannot, by itself, answer whether a person was close enough, oriented appropriately, or present while the billboard was working. Calling the traffic number an audience number discards the distinction between a place and a medium.
A useful thought experiment is to hold the traffic total constant while changing the billboard. Put one asset on a broad, unobstructed approach where most vehicles travel toward the face of the display. Put another behind a bend, partly screened by permanent structures, with a much narrower angle from the roadway. The roadway count may be identical, but the media opportunity is not. Now hold the asset constant and change the time period: daylight, darkness, a temporary closure, or a season with a different travel pattern. Again, the location total alone does not tell us whether the display was available in the same way. These examples do not provide a universal conversion factor; they show why the conversion needs asset- and time-specific evidence.
The same scene can contain several valid observations:
- vehicles crossed a defined roadway segment;
- people were present in the area around the asset;
- the people were within a Display Exposure Zone, the physical area in which the display could be seen or heard;
- the display was functioning and met a viewability condition while they were present;
- the people had some evidence-based likelihood of noticing or seeing the content;
- unique individuals met the further requirements for Audience.
These are related observations, but each answers a different question. The first asks who passed a place. The later questions ask what relationship those people had to a functioning medium and to its content. A lower-level measure can be useful without being defective. It simply has to keep its proper name.
The next level adds the asset. Under the MRC framework, Gross Impressions are potential ad exposures over a period when Presence exists in the Display Exposure Zone and the display is functional. The Display Exposure Zone is the physical area in which a person has an opportunity to see or hear a specific OOH source; its definition depends on factors such as the display’s size, the distance from it, and the viewing angle. The zone is therefore narrower and more asset-specific than a general location boundary.
“Functional” also has a concrete meaning. A digital display must be working as intended and able to display content. An analog billboard must be installed and properly maintained during the relevant period, subject to the qualifications for rotation, switching, or obstruction. A Gross Impression is thus stronger than a raw traffic count: it links presence to a functioning display in a defined area. But it remains the least restrictive impression level in the ladder. A person can be present in the zone while an obstruction affects what they can see, and Gross Impressions do not necessarily remove every such person.
Gross is not a synonym for unique reach. More than one impression can be counted for an individual when evidence shows that they entered the zone on separate occasions, and multiple ads may be counted during presence when the relevant conditions are met. This is why a gross total can exceed the number of distinct people. It answers a question about potential exposures across a period, not necessarily “how many different people saw this message?”
The distinction between “potential” and “unique” matters in a familiar commute. The same driver may pass the same billboard five mornings in one week. A gross measure may preserve those separate opportunities, while a reach measure would attempt to count the person once within the defined reporting period. Neither treatment is automatically correct for every decision. The report has to say which question it is answering. A campaign that values repeated opportunities may care about frequency; a campaign comparing unduplicated reach across media needs a defensible method for deciding when two observations belong to the same person. The first is a count of exposures, the second is a claim about identity and duplication. They should not be allowed to merge simply because both are displayed as large totals.
What viewability and audience add
OTS, or Opportunity-to-See Impressions, is also called Viewable Impressions in the MRC standard. It retains the prerequisites for Gross Impressions and adds a viewability condition. For OOH, the standard applies a condition that the full static posting or pixels be on the asset for one continuous second, or that video content be present for two continuous seconds. Logical or on-screen obstructions can negate the viewability condition; environmental obstructions are treated as a consideration for the later LTS and Audience levels. [MRC]
This is a crucial boundary: viewability is an opportunity condition, not a report that a person looked. It establishes that the content was present in a form that could be viewed. It does not tell us what the person was attending to, whether they faced the display, or whether they noticed the message. The standard explicitly distinguishes the availability of content from evidence that it was seen.
A duration rule also does not mean that every qualifying interval has the same human meaning. One person may move through the zone in a direct line with a clear view; another may meet the same time condition while facing across the street. The rule supplies a consistent minimum condition for the metric. It does not eliminate the need for the later model or evidence that handles direction, visual clutter, apparent size, and notice. This is why a viewability number can be more informative than traffic without being a measurement of attention.
Consider two people in the same exposure zone. One is walking toward the billboard with a clear line of sight. The other is standing in the same zone but is looking at a station entrance. If all the measurement system knows is their presence and the functioning display, both may contribute to a Gross measure. If the viewability condition is also satisfied, they may meet the OTS condition under the relevant method. That does not make their mental states identical. OTS is designed to avoid claiming more than the exposure opportunity supports.
LTS, or Likelihood-to-See Impressions, adds another refinement. It begins with the OTS conditions and applies an adjustment or determination based on some evidence of notice or being seen. The word “likelihood” is doing important work. It signals an evidence-based probability, not a direct observation of every person’s attention. The probability can vary with the display’s characteristics, the surrounding environment, traffic direction, illumination, distance, and the strength of the measurement technique. A model that is reasonable for a large, face-on roadside billboard may not transfer unchanged to a small, oblique screen in a crowded venue.
Audience is more stringent still. The MRC definition describes Audience as individuals estimated to be within the Display Exposure Zone who meet requirements for Presence, display functionality, viewability, and evidence of consumption, with additional criteria beyond LTS. “Consumption” here does not mean that the ad caused a purchase or changed someone’s behavior. It means that the measurement has stronger evidence of contact with the content. Audience is not an outcome metric, and it should not be presented as proof of conversion.
The move from LTS to Audience therefore requires more than a higher-sounding percentage. The measurement organization needs a defensible basis for its adjustment. The standard calls for attention to apparent size, viewing angle, environmental or visual obstruction, audio where relevant, exact placement, dynamic content, visual clutter, illumination and seasonality, exposure time, and the speed, direction, and mode of travel. It also calls for empirically supported thresholds and deterministic data that are representative, sufficiently covered, recent, granular, and subject to validation. The specific method may vary, but the burden is to show why it fits the asset and population being measured.
There is a useful difference between what is observed, what is inferred, and what is reported. A camera or manual count may observe presence. A geometric model may infer whether the person was within a valid zone. A calibrated adjustment may infer a likelihood of notice. The reporting system then assigns a label such as OTS, LTS, or Audience. Each step can be reasonable while still adding uncertainty. A reader should therefore ask not only “what did the sensor see?” but also “what transformation occurred before this number received its label?”
The ladder can be pictured as a sequence of added conditions:
The arrows should not be read as a promise that every traffic count can be converted into Audience by applying a standard multiplier. Each transition changes the claim. The first transition makes the media asset relevant. The second adds availability. The third adds a probability of notice. The fourth adds stronger evidence of consumption. A measurement program may be strong at one transition and weak at another.
Why the labels must stay attached
The most reliable way to compare OOH numbers is to keep the qualifying basis next to the number. “One million impressions” is incomplete unless the report explains whether these are Gross, OTS, LTS, or Audience Impressions; whether they refer to the display or a particular ad; whether they are gross or unique; and what time period, population, and exposure zone were used.
The same discipline applies to Reach and Frequency. Reach is a net or unduplicated count or percentage of a defined universe exposed within a defined period. Frequency describes the number of times that universe is typically exposed in the same period. A gross reach estimate, an OTS reach estimate, and an Audience reach estimate may all be mathematically coherent while representing different populations and qualification rules. The MRC standard recommends reporting Reach and Frequency on an Audience basis for cross-media comparison, while allowing alternate bases only when they are clearly labeled and Audience is also reported. [MRC]
This is also why the word “audience” cannot be used casually as a synonym for “people nearby.” In a billboard proposal, a traffic total may help answer whether a site has scale. A Gross Impression estimate may add the functioning asset and presence in its exposure zone. OTS may add the opportunity condition. LTS may provide a modeled likelihood of notice. Audience may support a more demanding claim if evidence of consumption and the additional qualifications are present. The labels are not a hierarchy of moral value; they are a hierarchy of what the measurement has established.
The hierarchy also protects against an opposite mistake: rejecting every lower-level measure because it is not Audience. Traffic and circulation remain useful when their purpose is stated. They can help size a location, identify variation, or provide a ceiling for later estimates. OTS can be useful when the decision concerns the availability of a display rather than conscious contact. The error is not using a lower-level measure. The error is using it while asking the reader to infer the higher-level claim.
For a reader evaluating a dashboard, the first diagnostic question is therefore simple: “What conditions did this number meet?” If the answer is only that people or vehicles were present near a billboard, the result is a traffic or circulation observation. If the answer includes a functioning display and exposure zone, it may support a Gross measure. If it includes the viewability condition, it may support OTS. If it adds evidence of notice or seen, it may support LTS. If it adds the further evidence and qualification required by the chosen standard, it may support Audience. The next challenge is to understand how those physical conditions are established for people who are moving through the scene.