When advertisers play Indians
How to Use Tracking Studies to Measure the Quality of Campaigns and Revamp Existing Ones
How tracking studies can measure the qualities of campaigns and turn existing ones aroundWolfgang J. Koschnick There is a remedy for all advertising ills. It works wonders when campaigns start to flounder, and even during the course of a campaign, it can still be used to turn the tide. At least, that's what market researchers promise when they praise their advertising tracking services. But do tracking studies really deliver what they promise?
In the Anglo-Saxon world, tracking or advertising tracking are nowadays forms of continuous observation of advertising effects in the course of a campaign. These are attempts to overcome the major dilemma of classic behaviorist advertising research. As is well known, the latter regards the advertising event as a black box. What goes on inside can only be deduced from the stimulus-response scheme. A triggering stimulus (advertising) in front of the black box leads to a reaction (advertising effect) afterwards.
Classical advertising research uses pretests and posttests: If something that had been measured before changed afterwards, the change must have occurred by means of advertising. As long as there was no doubt that this is how advertising works, doubts about this scheme did not arise in the first place.
In the meantime, however, we know that advertising does not work this way, and that is why this measurement method is highly inadequate. In practice, advertising tracking means that similar samples of target persons are questioned about a specific advertising campaign in several survey waves in order to find out whether anything has changed with the respondents since the campaign began, i.e. whether it has triggered any effects.
Market researchers promise to solve all problems with their tracking studies. This instrument can be used to constantly monitor the success of an advertising campaign as a whole and make ongoing corrections to the campaign.
In a way, this is like repairing the race car while the engine is still running, and without a pit stop either. Even if a campaign started off on the wrong foot, tracking could be used to correct mistakes in the ongoing campaign, and a boring advertising campaign would ultimately become a masterpiece of high-impact communications art.
But in reality, the grapes hang much higher. The all-important question is also the fundamental problem of advertising effectiveness research: How do you define and operationalize advertising effectiveness? Most of the commercial tracking offers on the market fail to answer this question.
No matter how many waves of surveys are passed over a target group: If the crucial questions about advertising impact are wrongly posed, poorly operationalized, and also misleadingly interpreted and analyzed, even a large number of survey waves at short intervals and even a large variety of questions will not help. This always produces only "findings": data without meat, numbers
without meaning, results without insight value.
Even the stroke of luck throws
many questions
The limits of the advertising tracking instrument are much narrower than its providers promise. One has to be very careful when interpreting the findings. However, this approach and a subtle restraint in evaluating survey data are not exactly among the outstanding characteristics of the advertising industry.
As a continuation of the not-so-good old black box models with more modern means, all tracking studies are posttests. They assume that the use of stimuli (the advertising) leads to reactions on the market (the advertising effect). And they measure these reactions sequentially.
You would only know what an advertising campaign does if you could find out what is really going on in the black box. However, indicators are available for this at best. And at best, they indicate how advertising might work. They are not the effect itself. Even the simplest conceivable case is more complicated than it appears at first glance: If sales of a product increase significantly following a campaign, that's heaven on earth for the advertisers. But even this stroke of luck raises many questions that the advertising industry should seriously address:
- Was it really the advertising campaign that triggered the market success? Would it have happened even without it? Every advertiser has the spontaneous answer to this: "Only an unworldly academic could ask such a question."
- Couldn't an even greater market success have been achieved with a slightly different campaign or a slightly stronger advertising pressure? Conversely, wasn't one's own advertising success impaired by a competitor's much higher advertising pressure?
- What is the long-term impact of the campaign, which usually matters more than short-term sales success?
These are all - incidental - questions that need to be asked if one is serious about accountability in advertising. WerbeWoche has dealt with this in detail in two previous issues. At least some aspects of these fundamental questions can be answered not too badly with the help of tracking studies. The goal is to go beyond the effects on sales and also to determine the psychological effects of an advertising campaign on consumers.
Investigation objectives must be operationalized
For this purpose, consumers are surveyed in campaigns accompanying surveys in several waves. Tracking studies are therefore continuous wave surveys conducted at regular intervals as posttests. The survey is carried out on a representative sample of mostly 200 to 500 changing persons by personal or telephone interviews.
As a rule, tracking tests are designed to answer the following questions: What absolute effect is achieved with the given advertising effort? How do the values examined react to changes in advertising expenditure? How do the values change over time? Are the observed changes one-off events or are they clear trends?
Advertising tracking is usually directed at the "relevant target group" of market communication. Normally, these are the users of the product category under investigation. If the campaign is for a candy bar, a sample of candy bar buyers is chosen. In saturated markets where market share is at stake, this is reasonable. But of course, in new, growing markets, it can also make sense to include in the sample people who are not yet using the new product.
Advertising objectives are unsuitable for a standardized survey
The all-important question now is: How are the research objectives operationalized in advertising tracking in order to identify the impact of specific campaigns? This is where, as a rule and in individual cases, the rabbit is in the hay. Because what is measured in the studies is not the advertising impact itself, but a series of indicators of advertising impact. This is quite unavoidable in standardized surveys, but not exactly mandatory either.
It would also be possible to measure only the advertising impact as such. However, this would require crystallizing the exact advertising objectives. And these vary from case to case and from campaign to campaign, so they are often not suitable for a standardized survey.
Each tracking model has its own peculiarities. But in general, the following indicators of advertising impact are measured as standard in advertising tracking studies:
- the recognition of a brand (brand awareness)
- the memory of brand advertising (Advertising Recall)
- the recollection of certain contents of brand advertising (copy recall)
- the recognition of the message of an advertising message (Advertising Awareness)
- the attitude toward the brand, or toward advertising a brand (Attitude toward Brand)
- the image of a brand (brand image)
- the liking of a brand
(Liking of Brand)
- the brand preference (Brand Preference)
- Purchase Intention
Which and how many of these and similar indicators are included in the study varies from institute to institute and also depends on the purpose of the study. It is also a question of money, because advertising tracking is expensive. With less than 250000 francs useful results are hardly to be won in a campaign. And which medium-sized or even somewhat larger company has this amount at its disposal for nothing more than the finding that its own campaign is not working or is working quite well?
Tracking has many pitfalls and traps that you need to know about
Even then, tracking is an adventure in which one can stumble from pitfall to pitfall. If indicators such as recall or awareness are surveyed, memory distortions occur: The majority of respondents claim they saw a certain ad on TV, even though they actually heard it on the radio or read it in an ad. No wonder advertising tracking is so popular with TV advertising marketers.
Another source of error arises from the usually narrow time frame. Even if it covers the entire duration of a campaign, it still neglects longer-term advertising effects that are important for building a brand. Long-term advertising effects are swept under the rug in tracking studies. This leads to a distortion: Transitory media systematically perform better than statuesque advertising media, which can still be busy building up contacts weeks and months after the conclusion of a campaign, even though no tracking study notices this yet.
This means that tracking studies are stabbing advertising in the back. Everything that is important about advertising and distinguishes it from sales and sales promotion, i.e., the long-term communication of a message and the lasting creation of a brand image, can no longer be measured by tracking studies because they are long since finished when these effects occur. Tracking studies are generally too short-winded to be able to measure what advertising is supposed to achieve overall.
The result of any tracking study is a comparison of the progress curves of the surveyed indicators, such as advertising recall, with those of advertising expenditure. But the implied causality of the relationship between advertising expenditure (equal to advertising spending) and advertising impact is highly problematic. This is because, of course, the impact of advertising is not solely dependent on advertising expenditure, but also on the use of all the other instruments in the marketing mix.
And quite naturally, qualitative components such as the creative implementation of an advertising concept often play a greater role than the amount of money invested. All in all, then, even the immediate findings from a tracking study are of limited value at best.
By contrast, the intensive experience of an institute, built up over many years of practical advertising tracking, is virtually invaluable. Year after year, they accumulate data and facts about advertising campaigns for different products, product categories, brands and advertisers. And from year to year the treasure of practical marketing experience grows. And the cumulative knowledge that is literally hidden in databases is often more valuable than any individual insight from a tracking study.
In the Anglo-Saxon world, tracking or advertising tracking are nowadays forms of continuous observation of advertising effects in the course of a campaign. These are attempts to overcome the major dilemma of classic behaviorist advertising research. As is well known, the latter regards the advertising event as a black box. What goes on inside can only be deduced from the stimulus-response scheme. A triggering stimulus (advertising) in front of the black box leads to a reaction (advertising effect) afterwards.
Classical advertising research uses pretests and posttests: If something that had been measured before changed afterwards, the change must have occurred by means of advertising. As long as there was no doubt that this is how advertising works, doubts about this scheme did not arise in the first place.
In the meantime, however, we know that advertising does not work this way, and that is why this measurement method is highly inadequate. In practice, advertising tracking means that similar samples of target persons are questioned about a specific advertising campaign in several survey waves in order to find out whether anything has changed with the respondents since the campaign began, i.e. whether it has triggered any effects.
Market researchers promise to solve all problems with their tracking studies. This instrument can be used to constantly monitor the success of an advertising campaign as a whole and make ongoing corrections to the campaign.
In a way, this is like repairing the race car while the engine is still running, and without a pit stop either. Even if a campaign started off on the wrong foot, tracking could be used to correct mistakes in the ongoing campaign, and a boring advertising campaign would ultimately become a masterpiece of high-impact communications art.
But in reality, the grapes hang much higher. The all-important question is also the fundamental problem of advertising effectiveness research: How do you define and operationalize advertising effectiveness? Most of the commercial tracking offers on the market fail to answer this question.
No matter how many waves of surveys are passed over a target group: If the crucial questions about advertising impact are wrongly posed, poorly operationalized, and also misleadingly interpreted and analyzed, even a large number of survey waves at short intervals and even a large variety of questions will not help. This always produces only "findings": data without meat, numbers
without meaning, results without insight value.
Even the stroke of luck throws
many questions
The limits of the advertising tracking instrument are much narrower than its providers promise. One has to be very careful when interpreting the findings. However, this approach and a subtle restraint in evaluating survey data are not exactly among the outstanding characteristics of the advertising industry.
As a continuation of the not-so-good old black box models with more modern means, all tracking studies are posttests. They assume that the use of stimuli (the advertising) leads to reactions on the market (the advertising effect). And they measure these reactions sequentially.
You would only know what an advertising campaign does if you could find out what is really going on in the black box. However, indicators are available for this at best. And at best, they indicate how advertising might work. They are not the effect itself. Even the simplest conceivable case is more complicated than it appears at first glance: If sales of a product increase significantly following a campaign, that's heaven on earth for the advertisers. But even this stroke of luck raises many questions that the advertising industry should seriously address:
- Was it really the advertising campaign that triggered the market success? Would it have happened even without it? Every advertiser has the spontaneous answer to this: "Only an unworldly academic could ask such a question."
- Couldn't an even greater market success have been achieved with a slightly different campaign or a slightly stronger advertising pressure? Conversely, wasn't one's own advertising success impaired by a competitor's much higher advertising pressure?
- What is the long-term impact of the campaign, which usually matters more than short-term sales success?
These are all - incidental - questions that need to be asked if one is serious about accountability in advertising. WerbeWoche has dealt with this in detail in two previous issues. At least some aspects of these fundamental questions can be answered not too badly with the help of tracking studies. The goal is to go beyond the effects on sales and also to determine the psychological effects of an advertising campaign on consumers.
Investigation objectives must be operationalized
For this purpose, consumers are surveyed in campaigns accompanying surveys in several waves. Tracking studies are therefore continuous wave surveys conducted at regular intervals as posttests. The survey is carried out on a representative sample of mostly 200 to 500 changing persons by personal or telephone interviews.
As a rule, tracking tests are designed to answer the following questions: What absolute effect is achieved with the given advertising effort? How do the values examined react to changes in advertising expenditure? How do the values change over time? Are the observed changes one-off events or are they clear trends?
Advertising tracking is usually directed at the "relevant target group" of market communication. Normally, these are the users of the product category under investigation. If the campaign is for a candy bar, a sample of candy bar buyers is chosen. In saturated markets where market share is at stake, this is reasonable. But of course, in new, growing markets, it can also make sense to include in the sample people who are not yet using the new product.
Advertising objectives are unsuitable for a standardized survey
The all-important question now is: How are the research objectives operationalized in advertising tracking in order to identify the impact of specific campaigns? This is where, as a rule and in individual cases, the rabbit is in the hay. Because what is measured in the studies is not the advertising impact itself, but a series of indicators of advertising impact. This is quite unavoidable in standardized surveys, but not exactly mandatory either.
It would also be possible to measure only the advertising impact as such. However, this would require crystallizing the exact advertising objectives. And these vary from case to case and from campaign to campaign, so they are often not suitable for a standardized survey.
Each tracking model has its own peculiarities. But in general, the following indicators of advertising impact are measured as standard in advertising tracking studies:
- the recognition of a brand (brand awareness)
- the memory of brand advertising (Advertising Recall)
- the recollection of certain contents of brand advertising (copy recall)
- the recognition of the message of an advertising message (Advertising Awareness)
- the attitude toward the brand, or toward advertising a brand (Attitude toward Brand)
- the image of a brand (brand image)
- the liking of a brand
(Liking of Brand)
- the brand preference (Brand Preference)
- Purchase Intention
Which and how many of these and similar indicators are included in the study varies from institute to institute and also depends on the purpose of the study. It is also a question of money, because advertising tracking is expensive. With less than 250000 francs useful results are hardly to be won in a campaign. And which medium-sized or even somewhat larger company has this amount at its disposal for nothing more than the finding that its own campaign is not working or is working quite well?
Tracking has many pitfalls and traps that you need to know about
Even then, tracking is an adventure in which one can stumble from pitfall to pitfall. If indicators such as recall or awareness are surveyed, memory distortions occur: The majority of respondents claim they saw a certain ad on TV, even though they actually heard it on the radio or read it in an ad. No wonder advertising tracking is so popular with TV advertising marketers.
Another source of error arises from the usually narrow time frame. Even if it covers the entire duration of a campaign, it still neglects longer-term advertising effects that are important for building a brand. Long-term advertising effects are swept under the rug in tracking studies. This leads to a distortion: Transitory media systematically perform better than statuesque advertising media, which can still be busy building up contacts weeks and months after the conclusion of a campaign, even though no tracking study notices this yet.
This means that tracking studies are stabbing advertising in the back. Everything that is important about advertising and distinguishes it from sales and sales promotion, i.e., the long-term communication of a message and the lasting creation of a brand image, can no longer be measured by tracking studies because they are long since finished when these effects occur. Tracking studies are generally too short-winded to be able to measure what advertising is supposed to achieve overall.
The result of any tracking study is a comparison of the progress curves of the surveyed indicators, such as advertising recall, with those of advertising expenditure. But the implied causality of the relationship between advertising expenditure (equal to advertising spending) and advertising impact is highly problematic. This is because, of course, the impact of advertising is not solely dependent on advertising expenditure, but also on the use of all the other instruments in the marketing mix.
And quite naturally, qualitative components such as the creative implementation of an advertising concept often play a greater role than the amount of money invested. All in all, then, even the immediate findings from a tracking study are of limited value at best.
By contrast, the intensive experience of an institute, built up over many years of practical advertising tracking, is virtually invaluable. Year after year, they accumulate data and facts about advertising campaigns for different products, product categories, brands and advertisers. And from year to year the treasure of practical marketing experience grows. And the cumulative knowledge that is literally hidden in databases is often more valuable than any individual insight from a tracking study.
