Management by clutter
In European poster research, there are no uniform models for measuring advertising impact
In European poster research, there are no uniform models for measuring advertising impactBy Wolfgang Koschnick Mobility, traffic estimation, and visibility. These three characteristics of the British Oscar poster research currently permit the most reliable statements about the qualities of outdoor advertising sites. Summarized observations have been replaced by individual site evaluations. What they are capable of achieving in detail is shown by the example from Great Britain.
Europe's outdoor advertising industry presents a checkered picture. In some countries, it is virtually marginalized. Leading poster nations are France and Belgium, by any measure. Outdoor advertisers in Belgium rake in a staggering 13 percent of total advertising revenue, and those in France a full 12 percent. In France, the newspaper press as a whole only accounts for 13.8 percent. The only other country in the world with a similar weighting of outdoor advertising is Japan, with 12.8 percent.
But time and again, there are countries that completely break ranks. In Switzerland, outdoor advertisers achieve a market share of 11.5 percent. This means that the Swiss outdoor advertising industry generates almost the same amount of revenue as television advertising. In most countries, the classic media are always somewhere on a similar level. But in outdoor advertising, there are almost breathtaking differences from country to country. Billboard advertising simply cannot be lumped together.
Moreover, only in Austria, the UK, Germany, Spain, Italy, Greece, the Netherlands, and - albeit only recently - in most Eastern European countries does the poster still play a somewhat significant role. In all other countries, outdoor advertising expenditures and the share of the total advertising pie are rather a quantité négligeable.
Thus, outdoor advertising in the sparsely populated countries of Scandinavia is of little consequence. And that is why it is at best of peripheral interest how poster research is organized in these countries. Mostly not at all or poorly. And is hardly worth mentioning.
Outdoor advertising research dances to a different tune
But how have Europe's leading poster nations organized their research? Are there commonalities or at least common trends? Is at least the big pack moving in the same direction, even if a few scattered horsemen are pushing to the rear? In fact, European - and with it global - outdoor advertising research has experienced some quantum leaps in recent decades and this year seems to be moving very gradually toward a hybrid model that, in one concrete form or another, finds acceptance in most European countries.
The poster is a complicated medium, and poster research is the high freestyle of media research. General advertising media research can be compared to the production of condoms: The same piece of rubber is always put over the same plunger. For every advertising medium, you need contact figures, frequencies, reach, users in the publication interval, and much more. And in principle, this works in the same way for all media. But it doesn't work for outdoor advertising. There are at least two reasons for this:
With "normal" advertising media, there are "normal" users (readers, TV viewers, radio listeners, moviegoers) who also have advertising contacts on the side. In the case of billboards, these users do not exist. The advertising medium "billboard" per se has no audience at all. No one uses a billboard if there is no poster on it.
The advertising medium is worth nothing to the advertiser if there is no advertising material on it. The opposite is true for "normal" advertising media. In poster research, therefore, the use of the advertising medium "billboard" does not provide any clues as to the use of the advertising medium "poster.
The "normal" advertising media are all the same: A daily newspaper is the same paper everywhere in its publication area. A magazine is the same everywhere, regardless of whether you read it in Basel, Zurich or Bern. The program of a radio or a television station reaches every listener and every viewer just as it does every other listener or viewer. The sameness of the advertising media is their overriding characteristic. Therefore, they are easy to measure and compare.
With the billboard, it's completely different. One poster site is located in the middle of the hustle and bustle of Zurich Airport. Hundreds of thousands pass by there every day. The other is in a small town where not much is going on, but thousands of people still pass by. And yet another stands in a remote village with its front facing the forest. The fox and the hare don't even see it when they say good night to each other.
In short, no two billboard sites are alike. The diversity of advertising media is the basic characteristic of outdoor advertising. Similarity is the characteristic of all other advertising media. Poster research must come to terms with this if it wants to do justice to the special characteristics of the poster as an advertising medium. Recently, four different models have come to play a role in the European discussions on the right path for poster research:
- the british postar model
- the dutch "Buitenreclame 2000
- the German PMA model
- the Austrian PHE model
From these most developed basic models, some of which differ greatly from each other, recent trends in outdoor advertising research can be identified.
Since the mid-1990s, the British Postar model has been regarded as a pioneering model. Postar is an acronym made up of the words "poster advertising research," cleverly using the consonance of the words "poster" and "postar" and also sounding like "poster star" in English. It is a continuation of the old Oscar model from the eighties with more modern means and is in the long tradition of British outdoor advertising research. Today, however, it is no longer even supported by the weight of outdoor advertising in the British Isles; because Great Britain has long since ceased to be one of the leading poster nations in Europe - as it was ten or twenty years ago.
The first model to set international standards also came from England. It was created soon after the end of the Second World War in the mid-1950s. Outdoor advertising had a different status then than it does today. Immediately after the war, there were more important things than advertising with posters. The struggle for survival dominated everyday life. Even as economic development progressed, outdoor advertising was more of a secondary matter. There were few billboards. Posters were not very common as an advertising medium.
Advertising in general was insignificant in the seller's markets of the 1950s and even in the 1960s: it was not demand that determined supply, but everything that was offered was in demand. As long as there was scarcity, advertising was meaningless, and outdoor advertising was even more so; because only a minority of consumers even owned an automobile or were even remotely as mobile as people are today.
The A-value was created when Europe was still in ruins
In this situation, it was necessary to find a model of poster research tailored to the conditions of the time, with the help of which advertisers could be given a rough guide so that they could assess the advertising suitability of different billboards in a broad and reasonably reliable way. In the process, it was possible to sweep under the table all the factors that are of central importance today for the advertising effectiveness of billboards: Neither the now exponentially increased mobility of the population nor the different chances of attention for different posting locations played an important role in assessing their effectiveness.
The decisive factor was that the model for poster research had to be simple and easy to use. And so the British market researcher Brian C. Copland came up with a model that was downright ingeniously simple by the standards of the time. It was usually called the "Copland model" after its inventor. The formula Copland had developed was called the A-value. And because the entire British advertising industry considered this value to be the pioneering poster research model par excellence, the Institute of Practitioners in Advertising (IPA) - the research arm of the Advertising Association (AA) - simply called it the "IPA formula.
When Brian Copland began preliminary work on the development of his model of poster research in 1949, Europe was still in ruins. This explains many things, including the ingenious simplicity of his A-value. European cities and communities were still far from the unheard-of levels of complexity, flexibility, and mobility that characterize them today. And even in 1956, when Copland was commissioned by the Institute of Practitioners in Advertising to conduct his groundbreaking "Newport Study," there was far less media diversity against which outdoor advertising had to compete than there is today. Television played virtually no role as a mass medium. In order to measure the traffic around billboard sites in a reasonably effective way, only a few data were needed.
In the situation at that time, the A-value was without any doubt the ultimate in poster research. As a measurement factor for contacts per capita of the population and poster site in a municipality, it simply takes into account the total number of contacts or passages in an occupancy period, the population of a location, and the number of poster sites. From a mathematical-statistical point of view, the average probability of passage is a function of the number of inhabitants in the locality.
To put it more simply, if you have the number of stop locations and the number of inhabitants, you automatically get the passage probability. The more inhabitants a place has, the smaller the value for the probability of passage. Basically, the only variable in the equation is the number of inhabitants, and it would then suffice to rank (or categorize) the cities and municipalities of a country according to their number of inhabitants in order to obtain a measure of the average probability of passage at the average stop locations that at least meets Stone Age standards.
Moreover, the approach underlying the A-value is virtually soaked in the philosophy of the arithmetic mean. Every differentiation in reality is rolled down to the mean of the average. Every stop has the same value, whether it is large or small, good or bad to see, positioned frontally or parallel.
The mobility of the population is not taken into account at all. The A-value does not take into account at all the phenomenon that today entire cities consist partly of purely residential areas and partly of purely commuter areas, and that a large part of the population travels daily from one city to the next or from one district to the other in order to get to work. For him, only the number of inhabitants counts.
The A-value had to give way to the Oscar and the Postar
In Great Britain, the realization grew in the course of the 1980s that Brian C. Copland's A value was no longer able to do justice to the growing differentiation and complexity of modern societies. And so people began to think about a modern model, after initially trying in vain to adapt the good old A-value to the requirements of modern times. This proved to be unfeasible.
In the UK, the A score has not been used at all since 1980. In France, there was a brief attempt at revival in a series of studies conducted by the Centre d'Etude de Support de Publicité (CESP) in various cities between 1983 and 1985. However, the CESP already used a further development of Copland's A-value after it became clear that the original value was no longer meaningful.
Today, variations of the A-value are still used in Switzerland, Belgium, Sweden and Poland. But there is no doubt that this is a discontinued model. In Sweden, for example, contact frequency and range are calculated according to a Copland model adapted to Swedish conditions by Gunnar Ekman.
And so, in 1981, an even more ingenious model was developed in Great Britain, which also has the advantage of being as differentiated as possible and fully meets the requirements of modern society: Oscar. The abbreviation stands for Outdoor Site Classification and Audience Research. In 1996, Oscar was further developed into Postar.
Simply put, Postar is something like a media analysis for the billboards of Great Britain (or any other country). As a continuous advertising media analysis, it offers both intermedia and intramedia comparability. As early as 1987, Oscar provided precise figures on the reach, contact frequency, and demographic composition of the target audience for virtually every one of the then total of 125,000 billboards in the British Isles. With Oscar and even more so with Postar, the advertising medium poster and its advertising medium billboard became just as transparent and manageable as television or any other print medium. A four-stage program led up to Postar:
- Total survey of all UK attack sites.
- For each stop, measures of its visibility were then developed as indicators of its contact quality.
- After that, they again determined contact opportunities (Opportunity to See [OTS]) for each individual attack site.
- In the fourth step - the quantum leap from Oscar to Postar - the reach of all individual billboards within the total population and within target groups and their frequency of contact with the respective billboard were determined.
With Oscar came the single point rating
The pioneering aspect for the future of poster research at Oscar was the individual site evaluation, which had not been known in this form before: Namely, the idea that contact with a posting site does not result from the passage alone, but is determined by characteristics of the site itself, its visibility, and its environment.
From this observational data, the study developed estimates of contact potential, reach, and contact frequency for each stop. At a later stage, the number of visibility factors was then reduced to seven.
Whatever one's view of the value of the model, the idea of single-item assessment has been around since Oscar. And it lives on today in most modern research models in one form or another. It is only the multiplicity of factors that has tended to cause headaches for researchers in recent times. The tendency toward multifactorization reached an absolute peak in the German FAW-Nielsen model, which attempted to narrow down contact with the help of no less than 37 different visibility factors. Today, there is a strong tendency worldwide to reduce visibility factors.
Three elements distinguish Postar
The basic idea behind Oscar was to determine contact chances, contact frequencies, and richness values based on qualifying location characteristics for passages. All people who passed a poster location within a certain distance were recorded as having a contact chance. This is where Postar later came in, providing a more realistic measure for assessing the impact of billboards of those people who are likely to actually notice the billboard and not just pass by it. Postar is characterized by three elements: a mobility study, traffic estimates, and a visibility study.
The purpose of the mobility study was to obtain data on the mobility behavior of adults using a representative sample. The survey ran in 140 locations across the UK. More than 7000 adults were interviewed, and details of all the roads a person walked or drove on were recorded.
The poster sites themselves are only fed into the analysis to calculate how many sites a respondent must have passed on journeys and walks. In this way, any number of new sites can be added to the database each year without having to conduct new interviews each time.
Traffic estimates: This data was used to correlate one-third of the billboard locations that Postar covers. For the remaining two-thirds, vehicle traffic counts were purchased from all communities that had these data. These counts were combined with 6175 separate locations from which potential modeling data already existed on a database. This covered over 10000 billboard sites - about ten percent of all sites. In addition, 9000 twelve-minute pedestrian counts were conducted at billboard sites across the country.
To model all the data material, neural networks were used, which were fed with the traffic and pedestrian counts until they were able to use the recorded information to generalize or estimate traffic at the previously unconsidered locations. The estimates were checked against real traffic counts and proved to be very accurate: for an average campaign with a thousand poster sites, the deviation was only ±0.6 percent.
The visibility study: In the next step, the data on passage was refined so that it was possible to calculate how many people were likely to have seen the poster. The poster visibility experiments were conducted by the Psychology Department at Birkbeck College, University of London. In the process, 50 test subjects who believed they were taking part in a road safety experiment were confronted with posters in a normal street scenario.
To determine the exact viewing behavior of subjects, Paul Barber of Birkbeck College used computer-developed still images of stop surfaces and a series of predetermined distractors. These were shown to test subjects and their viewing behavior was tracked. Their eye movements were analyzed with the help of infrared measuring technology, so that it could be determined who, when, where had perceived a poster.
It was found that fixations in gaze behavior - i.e. looking closely - were determined by five factors: the size and the angle of the surface to the observer (actually two factors); "eccentricity": this is the angle between the direction of travel and the direction one would need to see the surface fully. So, in a sense, this is the deviation from full visibility; illumination ("illumination"); visual clutter: the totality of factors (traffic signs, shop windows, environmental complexity, situational complexity, etc.) that distract the viewer from the billboard; duration of passage.
On this basis, it was possible to design a visibility model applicable to any known billboard location. It provides an estimate of the actual viewers or the duration of billboard viewing if the number of people passing by is known. The visibility experiment allows an estimate of effective contacts. Postar thus makes it possible to estimate the percentage of those target people who actually see a poster: the reach.
Postar's approach makes sense, but it is not comprehensible to everyone. The overall model is far too complicated. This is one of many reasons why enthusiasm for the model is waning in the British Isles. Postar has fallen into a strange state of limbo, where no one can really say at the moment whether it is going forward or backward again.
New approaches in poster researchWith this article, WerbeWoche begins a new series of articles on the state of current poster research in Europe. After this first part, which describes the basic problems of the outdoor advertising industry, the second article describes and analyzes the advantages and disadvantages of newer approaches in research, in particular the Dutch, German, and Austrian research models.
Outdoor advertising by expenditure in million euros 1999
France 1089
Great Britain 954
Germany 682
Italy 360
Switzerland 294
Spain 217
Belgium 159
Greece 154
Netherlands 130
Poland 126
Austria 119
Sweden 86
Ireland 48.1
Hungary 43.6
Denmark 42
Portugal 32
Finland 32
Norway 31
Slovakia 13.5
Luxembourg 1.6
Czech Republic 1.5
Romania 0.5
Source: IP Status Report 2000 Advertising and the Economy
The criteria established by OSCAR:- Road category of the stopping place: Here, the places were divided according to the highly visible road type.
- Location: Nine different location categories were differentiated, again using detailed definitions.
- Traffic flow: A distinction was made according to whether the traffic flow at the stop was one-way or two-way, or whether it was in a pedestrian zone with no automobile traffic.
- Number of lanes
- Speed limits
- Proximity of traffic lights within 100 meters of the stop location.
- Access routes: positions from which vehicles or pedestrians approach the impact site.
- Proximity to the city center: A distinction is made between five gradations.
- Proximity to the nearest shopping center: five gradations.
- Visibility for stores: This specifies whether the posting location can be seen from a store and what category of store it is (twelve categories are given).
- Special Features: Peculiarities (Special Features) of the stop location are recorded according to 21 predefined categories.
- Stop type: Twelve basic types are distinguished.
- Visibility: information on the visibility factors observed by field workers and recorded in a site sketch they prepared for each impact site.
- Length of Visibility: The furthest distance from which a poster can still be clearly seen at the respective posting location.
- Angle of Vision: The angle at which the point of impact is to the road.
- Deflection from Natural Line of Sight: The deviation from the natural line of sight - this is the estimate of the angle by which a driver must turn his head from his normal visual position while driving in order to see a stop.
- Presence of visual obstructions: Information about lampposts, traffic signs, bus stops, and the like that - in the estimation of the field workers used in the study - interfere with or even obstruct the clear view of passing road users to the impact site.
- Height of Panel from the Ground: Height above the ground.
- Illumination: illumination of the stop.
- Competition: The total number of billboards placed on a billboard site that together vie for the attention of passersby.
Europe's outdoor advertising industry presents a checkered picture. In some countries, it is virtually marginalized. Leading poster nations are France and Belgium, by any measure. Outdoor advertisers in Belgium rake in a staggering 13 percent of total advertising revenue, and those in France a full 12 percent. In France, the newspaper press as a whole only accounts for 13.8 percent. The only other country in the world with a similar weighting of outdoor advertising is Japan, with 12.8 percent.
But time and again, there are countries that completely break ranks. In Switzerland, outdoor advertisers achieve a market share of 11.5 percent. This means that the Swiss outdoor advertising industry generates almost the same amount of revenue as television advertising. In most countries, the classic media are always somewhere on a similar level. But in outdoor advertising, there are almost breathtaking differences from country to country. Billboard advertising simply cannot be lumped together.
Moreover, only in Austria, the UK, Germany, Spain, Italy, Greece, the Netherlands, and - albeit only recently - in most Eastern European countries does the poster still play a somewhat significant role. In all other countries, outdoor advertising expenditures and the share of the total advertising pie are rather a quantité négligeable.
Thus, outdoor advertising in the sparsely populated countries of Scandinavia is of little consequence. And that is why it is at best of peripheral interest how poster research is organized in these countries. Mostly not at all or poorly. And is hardly worth mentioning.
Outdoor advertising research dances to a different tune
But how have Europe's leading poster nations organized their research? Are there commonalities or at least common trends? Is at least the big pack moving in the same direction, even if a few scattered horsemen are pushing to the rear? In fact, European - and with it global - outdoor advertising research has experienced some quantum leaps in recent decades and this year seems to be moving very gradually toward a hybrid model that, in one concrete form or another, finds acceptance in most European countries.
The poster is a complicated medium, and poster research is the high freestyle of media research. General advertising media research can be compared to the production of condoms: The same piece of rubber is always put over the same plunger. For every advertising medium, you need contact figures, frequencies, reach, users in the publication interval, and much more. And in principle, this works in the same way for all media. But it doesn't work for outdoor advertising. There are at least two reasons for this:
With "normal" advertising media, there are "normal" users (readers, TV viewers, radio listeners, moviegoers) who also have advertising contacts on the side. In the case of billboards, these users do not exist. The advertising medium "billboard" per se has no audience at all. No one uses a billboard if there is no poster on it.
The advertising medium is worth nothing to the advertiser if there is no advertising material on it. The opposite is true for "normal" advertising media. In poster research, therefore, the use of the advertising medium "billboard" does not provide any clues as to the use of the advertising medium "poster.
The "normal" advertising media are all the same: A daily newspaper is the same paper everywhere in its publication area. A magazine is the same everywhere, regardless of whether you read it in Basel, Zurich or Bern. The program of a radio or a television station reaches every listener and every viewer just as it does every other listener or viewer. The sameness of the advertising media is their overriding characteristic. Therefore, they are easy to measure and compare.
With the billboard, it's completely different. One poster site is located in the middle of the hustle and bustle of Zurich Airport. Hundreds of thousands pass by there every day. The other is in a small town where not much is going on, but thousands of people still pass by. And yet another stands in a remote village with its front facing the forest. The fox and the hare don't even see it when they say good night to each other.
In short, no two billboard sites are alike. The diversity of advertising media is the basic characteristic of outdoor advertising. Similarity is the characteristic of all other advertising media. Poster research must come to terms with this if it wants to do justice to the special characteristics of the poster as an advertising medium. Recently, four different models have come to play a role in the European discussions on the right path for poster research:
- the british postar model
- the dutch "Buitenreclame 2000
- the German PMA model
- the Austrian PHE model
From these most developed basic models, some of which differ greatly from each other, recent trends in outdoor advertising research can be identified.
Since the mid-1990s, the British Postar model has been regarded as a pioneering model. Postar is an acronym made up of the words "poster advertising research," cleverly using the consonance of the words "poster" and "postar" and also sounding like "poster star" in English. It is a continuation of the old Oscar model from the eighties with more modern means and is in the long tradition of British outdoor advertising research. Today, however, it is no longer even supported by the weight of outdoor advertising in the British Isles; because Great Britain has long since ceased to be one of the leading poster nations in Europe - as it was ten or twenty years ago.
The first model to set international standards also came from England. It was created soon after the end of the Second World War in the mid-1950s. Outdoor advertising had a different status then than it does today. Immediately after the war, there were more important things than advertising with posters. The struggle for survival dominated everyday life. Even as economic development progressed, outdoor advertising was more of a secondary matter. There were few billboards. Posters were not very common as an advertising medium.
Advertising in general was insignificant in the seller's markets of the 1950s and even in the 1960s: it was not demand that determined supply, but everything that was offered was in demand. As long as there was scarcity, advertising was meaningless, and outdoor advertising was even more so; because only a minority of consumers even owned an automobile or were even remotely as mobile as people are today.
The A-value was created when Europe was still in ruins
In this situation, it was necessary to find a model of poster research tailored to the conditions of the time, with the help of which advertisers could be given a rough guide so that they could assess the advertising suitability of different billboards in a broad and reasonably reliable way. In the process, it was possible to sweep under the table all the factors that are of central importance today for the advertising effectiveness of billboards: Neither the now exponentially increased mobility of the population nor the different chances of attention for different posting locations played an important role in assessing their effectiveness.
The decisive factor was that the model for poster research had to be simple and easy to use. And so the British market researcher Brian C. Copland came up with a model that was downright ingeniously simple by the standards of the time. It was usually called the "Copland model" after its inventor. The formula Copland had developed was called the A-value. And because the entire British advertising industry considered this value to be the pioneering poster research model par excellence, the Institute of Practitioners in Advertising (IPA) - the research arm of the Advertising Association (AA) - simply called it the "IPA formula.
When Brian Copland began preliminary work on the development of his model of poster research in 1949, Europe was still in ruins. This explains many things, including the ingenious simplicity of his A-value. European cities and communities were still far from the unheard-of levels of complexity, flexibility, and mobility that characterize them today. And even in 1956, when Copland was commissioned by the Institute of Practitioners in Advertising to conduct his groundbreaking "Newport Study," there was far less media diversity against which outdoor advertising had to compete than there is today. Television played virtually no role as a mass medium. In order to measure the traffic around billboard sites in a reasonably effective way, only a few data were needed.
In the situation at that time, the A-value was without any doubt the ultimate in poster research. As a measurement factor for contacts per capita of the population and poster site in a municipality, it simply takes into account the total number of contacts or passages in an occupancy period, the population of a location, and the number of poster sites. From a mathematical-statistical point of view, the average probability of passage is a function of the number of inhabitants in the locality.
To put it more simply, if you have the number of stop locations and the number of inhabitants, you automatically get the passage probability. The more inhabitants a place has, the smaller the value for the probability of passage. Basically, the only variable in the equation is the number of inhabitants, and it would then suffice to rank (or categorize) the cities and municipalities of a country according to their number of inhabitants in order to obtain a measure of the average probability of passage at the average stop locations that at least meets Stone Age standards.
Moreover, the approach underlying the A-value is virtually soaked in the philosophy of the arithmetic mean. Every differentiation in reality is rolled down to the mean of the average. Every stop has the same value, whether it is large or small, good or bad to see, positioned frontally or parallel.
The mobility of the population is not taken into account at all. The A-value does not take into account at all the phenomenon that today entire cities consist partly of purely residential areas and partly of purely commuter areas, and that a large part of the population travels daily from one city to the next or from one district to the other in order to get to work. For him, only the number of inhabitants counts.
The A-value had to give way to the Oscar and the Postar
In Great Britain, the realization grew in the course of the 1980s that Brian C. Copland's A value was no longer able to do justice to the growing differentiation and complexity of modern societies. And so people began to think about a modern model, after initially trying in vain to adapt the good old A-value to the requirements of modern times. This proved to be unfeasible.
In the UK, the A score has not been used at all since 1980. In France, there was a brief attempt at revival in a series of studies conducted by the Centre d'Etude de Support de Publicité (CESP) in various cities between 1983 and 1985. However, the CESP already used a further development of Copland's A-value after it became clear that the original value was no longer meaningful.
Today, variations of the A-value are still used in Switzerland, Belgium, Sweden and Poland. But there is no doubt that this is a discontinued model. In Sweden, for example, contact frequency and range are calculated according to a Copland model adapted to Swedish conditions by Gunnar Ekman.
And so, in 1981, an even more ingenious model was developed in Great Britain, which also has the advantage of being as differentiated as possible and fully meets the requirements of modern society: Oscar. The abbreviation stands for Outdoor Site Classification and Audience Research. In 1996, Oscar was further developed into Postar.
Simply put, Postar is something like a media analysis for the billboards of Great Britain (or any other country). As a continuous advertising media analysis, it offers both intermedia and intramedia comparability. As early as 1987, Oscar provided precise figures on the reach, contact frequency, and demographic composition of the target audience for virtually every one of the then total of 125,000 billboards in the British Isles. With Oscar and even more so with Postar, the advertising medium poster and its advertising medium billboard became just as transparent and manageable as television or any other print medium. A four-stage program led up to Postar:
- Total survey of all UK attack sites.
- For each stop, measures of its visibility were then developed as indicators of its contact quality.
- After that, they again determined contact opportunities (Opportunity to See [OTS]) for each individual attack site.
- In the fourth step - the quantum leap from Oscar to Postar - the reach of all individual billboards within the total population and within target groups and their frequency of contact with the respective billboard were determined.
With Oscar came the single point rating
The pioneering aspect for the future of poster research at Oscar was the individual site evaluation, which had not been known in this form before: Namely, the idea that contact with a posting site does not result from the passage alone, but is determined by characteristics of the site itself, its visibility, and its environment.
From this observational data, the study developed estimates of contact potential, reach, and contact frequency for each stop. At a later stage, the number of visibility factors was then reduced to seven.
Whatever one's view of the value of the model, the idea of single-item assessment has been around since Oscar. And it lives on today in most modern research models in one form or another. It is only the multiplicity of factors that has tended to cause headaches for researchers in recent times. The tendency toward multifactorization reached an absolute peak in the German FAW-Nielsen model, which attempted to narrow down contact with the help of no less than 37 different visibility factors. Today, there is a strong tendency worldwide to reduce visibility factors.
Three elements distinguish Postar
The basic idea behind Oscar was to determine contact chances, contact frequencies, and richness values based on qualifying location characteristics for passages. All people who passed a poster location within a certain distance were recorded as having a contact chance. This is where Postar later came in, providing a more realistic measure for assessing the impact of billboards of those people who are likely to actually notice the billboard and not just pass by it. Postar is characterized by three elements: a mobility study, traffic estimates, and a visibility study.
The purpose of the mobility study was to obtain data on the mobility behavior of adults using a representative sample. The survey ran in 140 locations across the UK. More than 7000 adults were interviewed, and details of all the roads a person walked or drove on were recorded.
The poster sites themselves are only fed into the analysis to calculate how many sites a respondent must have passed on journeys and walks. In this way, any number of new sites can be added to the database each year without having to conduct new interviews each time.
Traffic estimates: This data was used to correlate one-third of the billboard locations that Postar covers. For the remaining two-thirds, vehicle traffic counts were purchased from all communities that had these data. These counts were combined with 6175 separate locations from which potential modeling data already existed on a database. This covered over 10000 billboard sites - about ten percent of all sites. In addition, 9000 twelve-minute pedestrian counts were conducted at billboard sites across the country.
To model all the data material, neural networks were used, which were fed with the traffic and pedestrian counts until they were able to use the recorded information to generalize or estimate traffic at the previously unconsidered locations. The estimates were checked against real traffic counts and proved to be very accurate: for an average campaign with a thousand poster sites, the deviation was only ±0.6 percent.
The visibility study: In the next step, the data on passage was refined so that it was possible to calculate how many people were likely to have seen the poster. The poster visibility experiments were conducted by the Psychology Department at Birkbeck College, University of London. In the process, 50 test subjects who believed they were taking part in a road safety experiment were confronted with posters in a normal street scenario.
To determine the exact viewing behavior of subjects, Paul Barber of Birkbeck College used computer-developed still images of stop surfaces and a series of predetermined distractors. These were shown to test subjects and their viewing behavior was tracked. Their eye movements were analyzed with the help of infrared measuring technology, so that it could be determined who, when, where had perceived a poster.
It was found that fixations in gaze behavior - i.e. looking closely - were determined by five factors: the size and the angle of the surface to the observer (actually two factors); "eccentricity": this is the angle between the direction of travel and the direction one would need to see the surface fully. So, in a sense, this is the deviation from full visibility; illumination ("illumination"); visual clutter: the totality of factors (traffic signs, shop windows, environmental complexity, situational complexity, etc.) that distract the viewer from the billboard; duration of passage.
On this basis, it was possible to design a visibility model applicable to any known billboard location. It provides an estimate of the actual viewers or the duration of billboard viewing if the number of people passing by is known. The visibility experiment allows an estimate of effective contacts. Postar thus makes it possible to estimate the percentage of those target people who actually see a poster: the reach.
Postar's approach makes sense, but it is not comprehensible to everyone. The overall model is far too complicated. This is one of many reasons why enthusiasm for the model is waning in the British Isles. Postar has fallen into a strange state of limbo, where no one can really say at the moment whether it is going forward or backward again.
New approaches in poster researchWith this article, WerbeWoche begins a new series of articles on the state of current poster research in Europe. After this first part, which describes the basic problems of the outdoor advertising industry, the second article describes and analyzes the advantages and disadvantages of newer approaches in research, in particular the Dutch, German, and Austrian research models.
Outdoor advertising by expenditure in million euros 1999
France 1089
Great Britain 954
Germany 682
Italy 360
Switzerland 294
Spain 217
Belgium 159
Greece 154
Netherlands 130
Poland 126
Austria 119
Sweden 86
Ireland 48.1
Hungary 43.6
Denmark 42
Portugal 32
Finland 32
Norway 31
Slovakia 13.5
Luxembourg 1.6
Czech Republic 1.5
Romania 0.5
Source: IP Status Report 2000 Advertising and the Economy
The criteria established by OSCAR:- Road category of the stopping place: Here, the places were divided according to the highly visible road type.
- Location: Nine different location categories were differentiated, again using detailed definitions.
- Traffic flow: A distinction was made according to whether the traffic flow at the stop was one-way or two-way, or whether it was in a pedestrian zone with no automobile traffic.
- Number of lanes
- Speed limits
- Proximity of traffic lights within 100 meters of the stop location.
- Access routes: positions from which vehicles or pedestrians approach the impact site.
- Proximity to the city center: A distinction is made between five gradations.
- Proximity to the nearest shopping center: five gradations.
- Visibility for stores: This specifies whether the posting location can be seen from a store and what category of store it is (twelve categories are given).
- Special Features: Peculiarities (Special Features) of the stop location are recorded according to 21 predefined categories.
- Stop type: Twelve basic types are distinguished.
- Visibility: information on the visibility factors observed by field workers and recorded in a site sketch they prepared for each impact site.
- Length of Visibility: The furthest distance from which a poster can still be clearly seen at the respective posting location.
- Angle of Vision: The angle at which the point of impact is to the road.
- Deflection from Natural Line of Sight: The deviation from the natural line of sight - this is the estimate of the angle by which a driver must turn his head from his normal visual position while driving in order to see a stop.
- Presence of visual obstructions: Information about lampposts, traffic signs, bus stops, and the like that - in the estimation of the field workers used in the study - interfere with or even obstruct the clear view of passing road users to the impact site.
- Height of Panel from the Ground: Height above the ground.
- Illumination: illumination of the stop.
- Competition: The total number of billboards placed on a billboard site that together vie for the attention of passersby.
