A study done by the Centre for Media Studies emphasizes the phenomenal growth in advertising that has taken place in India in recent years. It observes: “By any count 1986-88 years will go down as a turning point in the Indian advertising industry.” A number of factors have contributed to this growth. First, advertising expenditure has crossed the Rs.1000 crore mark. Second, advertising through TV has become a force to reckon with. Third, consolidation and realignment of advertising agencies has taken place. Fourth, newspapers and magazines have to woo advertisers and advertising agencies. Fifth, advertising expenditure is no longer questioned following the new buoyancy in the market.
Another study – shows that five consumer megatrends have swept across the urban landscape in India. These are: the new individualism, the growing multi-culture orientation, the search for status a harder look at brand values, and the new aspirations among the consumers. These magatrends have emerged as a result of four forces of change in the Indian society. These forces are: more purchasing power with consumers, availability of wide variety of products and brands, availability of greater information about products and brands, and changes in the socio-economic environment in the country. In view of these developments, advertising has assumed more importance in recent years. This rising trend will continue on account of a number of factors such as increase in urbanisation, expansion of education, increase in per capita income, etc.
This would mean that companies will give far more attention to their advertising efforts, use more sophisticated and improved advertising techniques and be more concerned about knowing the impact of their advertising efforts.
Need for Advertising Research: The role of research in advertising can be seen in various stages of advertising planning. First, what should be the objectives of advertising? Unless the objectives are clear, advertising cannot be useful. Research would enable the company to be clear in its objectives of advertising. Second, advertising research should be used for developing a strategy for marketing the product in question. Further, the selection of the target audience can be facilitated by advertising research. In this connection, mere demographic classification will not be adequate. It may be equally necessary to pay attention to product usage behaviour. Another aspect where research can be useful is the selection of message that an advertisement should carry and through what media it should be conveyed. This is a major area of advertising research. Finally, research has to concern itself with the evaluation of advertising in order to find out whether the expenditure on advertising has been justified or not. If not, the reasons should be ascertained so that in improvement in advertising can be made in the future. In short, research can be instrumental in increasing the efficiency of advertising as a result of which the pay-off from advertising expenditure will increase.
Media Research: The main issues in media research are: (a) How to choose amongst media types – television, radio and newspapers? (b) How to decide on a specific insert within a particular type of media, say, television? In order to decide on these two issues, it is necessary to have some data. The Advertising Research Foundation (ARF) suggested the following type of data for this purpose:
- Media vehicle distribution: the circulation number for a magazine or newspaper or the number of television or radio sets available to carry the advertising.
- Media vehicle audience: the number of people exposed to the media vehicle. This would be larger than the number in (1) above as more than one person reads the same newspaper/ magazine or watches on the same television set.
- Advertising exposure: the number of people exposed to a specific advertisement in the media vehicle. This number would be less than the number in (2) above as all those who are exposed to a newspaper/magazine may not notice a particular advertisement.
- Advertising perception: the number of people who perceived the advertisement in question. This number would be less than that in (3) above. The people may perceive an advertisement because of several factors such as its large size, use of attractive colours or its positioning in the media vehicle or on account of the product involved.
- Advertising communication: the number of people who comprehend specific things about the advertising. This number would obviously be less than the number of people who perceived.
- Sales response: the number of people who buy the product in question as a result of advertising. This number would be far less than that in (5) above.
It may be noted that of these six categories, there is an interaction of media and message in the last four categories. This makes it difficult to obtain the numbers for media alone in respect of these categories. Accordingly, media vehicle data are generally obtained for the first two categories, viz. media vehicle distribution and media vehicle audience.
Media Audiences: Media research comprises, inter alia, the measurement of the size and break-up of individual vehicle audience. We will discuss this under two heads – print media and radio and television.
Print Media: Prior to the reports of the Audit Bureau of Circulation (ABC), the measures regarding the individual vehicle audiences were those which the media themselves claimed. Such measures were rather inflated as any individual media vehicle would suggest that its circulation is far and wide. Since the ABC’s reports are now available, these inflated measures have ceased. The ABC compiles its report which gives the size of circulation of a newspaper/magazine on the basis of certified audits.
This information while being useful, it’s not sufficient. It is difficult to estimate precisely the size of audience for a particular publication. The data collected by merely asking respondents as to whether they have looked at a particular copy would be unreliable. This is because some respondents may regard reading a particular magazine as a status symbol and hence they may report exaggerated readership. Likewise, reading of some magazines may be regarded as below one’s status and hence their readership may be reported to be much less than in reality.
Another important aspect in determining the audience size for print media is the extent of duplication between magazines. This is because readership of three or more magazines among respondents is quite common. But the data on readership seldom give the extent of duplication. The problem is how to get the size of ‘unduplicated’ audience. A detailed study is determine the duplication among a large number of magazines would obviously be very expensive, not to mention the time it would involve.
Radio and Television: There are four methods to measure the size of the audience for any radio and television programme. These methods are discussed briefly below.
Coincidental Method: First of all, a sample of households having a telephone is selected. This is followed by an inquiry on telephone as to whether a particular programme on radio is being listened or being watched on television. Other information such as the name of the sponsor and the product being advertised is also collected.
The main advantages of this method are that it is quick and economical. It has some limitations though. First, the method has to be confined to only those households which own telephones. In a country like India, a large number of households do not have telephones and hence they have to be excluded. Second, since the enquiry has to be conducted while a particular programme is in progress only a limited number of households can be contacted in this short duration. Finally, it is extremely difficult to undertake an enquiry with respect to late night programmes.
Roster Recall: As the same names implies, a roster or list of programmes is used to facilitate respondents to recall what programmes were listened to or watched. Respondents are contacted personally by interviewers.
This method has some major limitations. First, the responses are dependent upon memory. Second, depending on the status or popularity or otherwise of a particular programme, respondents may give their replies regardless of whether they have listened to a programme (or seen it). Third, the method is unable to provide information on a continuing basis. Finally, it is not possible to estimate duplication in the audience as respondents are approached for programmes within short estimate duplication in the audience as respondents are approached for programme within a short time period. It is possible to estimate the number of persons who watch both programmes, falling within the same time span on which respondents are being contacted.
Diary Method: As the name implies, this method uses a diary for estimating the number of persons listening to or watching different programmes. A diary, especially designed for this purpose, is issued to respondents who have agreed to furnish the desired information. Each respondent records his radio listening or television viewing, along with personal data such as age and sex in this diary. If respondents accurately record their radio listening or television viewing, this method would give accurate and complete information, eliminating the errors that may arise due to memory lapse and interviewer bias. Further, it is cheaper than other methods involving personal interviewing and recall.
However, in practice, one may find that respondents are not so careful in listing the programmes listened to or viewed by them. Besides, there is a lack of continuity in the flow of information as the diary method is unable to provide the estimate of an audience, say, minute-by-minute. Apart from this, some respondents in the panel may stop giving the information sought or move to another address. In such a case, how far the panel will remain representative of the population is a moot question.
The Audimeter: As Audimeter is an electronic machine attached to a television set. As soon as the television set is turned on, the machine records it on the tape. In advanced countries, this method is frequently used. The audimeter ensures a continuous flow of information which is not possible in any of the earlier methods. This is its main advantage. Another advantage of this method is that there is complete objectivity in the information thus collected. Moreover it is possible to have a cumulative estimate of the audience since the audimeter sample will be almost the same from month to month. The method suffers from some limitations as well. First, turning a set on does not necessarily mean that the programme in question is being watched. Second, the method cannot indicate as to who is watching a programme.
Copy Testing: Another important area in advertising research is copy testing. The word ‘copy’ is used to denote an entire advertisement, including the message, pictures, colours, etc, regardless of the medium in which the advertisement has appeared. As Shirley Young says:
“Copy testing is troublesome for almost everybody. Most advertisers and agencies have a checkered history of its use and often differ on both what and how to measure. This continued dissatisfaction, with copy testing has led to a never ending search for the next new technique to embrace.”
Methods of copy testing can be divided into two categories, viz. ‘before’ tests and ‘after’ tests. The former category includes all those tests that are used in ascertaining the suitability or otherwise of an advertisement before it is finally released. Their purpose is to effect improvements in the copy or advertisement. The latter category includes tests to measure the effectiveness of an advertisement after it has been formally released. Despite this distinction, at times the difference between the two types of method gets blurred. This is because some ‘before’ methods require that an advertisement should be run one or two media.
‘Before’ Methods: A number of pre-test methods are used for copy testing. In this section, we will discuss the following methods: (i) consumer jury, (ii) portfolio tests, (iii) rating scales, (iv) physiological methods, (v) dummy advertising vehicles, (vi) on-the-air tests, (vii) inquiries, and (viii) laboratory testing.
Consumer Jury: In this method, a sizeable number of consumers from the target audience are shown a set of rough and unfinished advertisements. With respect to these advertisements, they are asked such questions as: Which copy would you prefer to read? Which one would induce you to buy the product? Which headline is the most interesting? Paired comparisons or ranking may be used by respondents. The assumption in this method is that at least one of the advertisements shown will be liked by them.
Rating Scales: This method involves the use of certain standards against which a copy is tested. They copy is rated on the basis of scale values. As a result, a numerical score is obtained. It may be added that weights may be assigned to different factors or items on the basis of which a copy is to be tested, depending on their relative importance or relevance. This method is generally used by professional advertising agencies which are able to ‘rate’ advertisements without any difficulty.
This method has one major advantage as it provides a list of factors against which a copy is to be tested. However, there are certain limitations. First, the problem is how eights are to be assigned to different items. Second, different respondents will rate the items differently. It is difficult to say who is right in his rating. Third, an overall high score of a copy does not necessarily mean a superior copy. This is because that copy might have scored high ratings with respect to several items and low ratings with respect to only a few items. It is these few items which may be extremely relevant in judging the copy.
Portfolio Tests: According to this method, a number of alternative copies that are to be tested are placed in a portfolio. At times, the copies are placed in dummy copies of magazines or newspapers. Respondents are given the portfolio and asked to go through it. After they have done so they are asked to recall the copies from memory. Such a recall may be either unaided or aided. The interviewer may facilitate recall by asking about specific advertisements. The interviewer may further ask the respondent to recall the advertisement as much as possible. The extent of recall will indicate the strength of the copy.
Psychological Tests: This method uses a number of psychological techniques to find out the reactions of respondents to a given advertisement. Techniques such as word association, sentence completion, depth interviewing and story telling are used by trained psychological to find our what respondent see in a given advertisement and the influence it has on them. As it is extremely difficult to interpret the information obtained on the basis of psychological tests, only trained persons should be appointed to carry out this test. In view of this, only a small sample can be used for such tests.
Laboratory Testing: This method uses mechanical devices to measure the respondent’s psychological responses to a given advertisement. The commonly used tests are the galvanic skin response and the eye movement. As regards the former, a device is used to measure changes in the amount of perspiration in the hands. This may be taken as a measure of emotional change as a response to an advertisement. However, the test is unable to indicate whether such an emotional change is favourable or unfavourable to an advertisement. In the latter test an eye camera registers the continuous movement of the eye as it reads an advertisement. However, the results obtained from the an eye camera are difficult to interpret. For example, if the eye was fixed on a certain point could it be interpreted that the respondent was interested in the advertisement or that he was confused?
Inquiry Tests: Some advertisements may invite several inquiries from the readers about a given product or service. However, it is questionable whether a large number of inquiries can be regarded as a good yardstick for a successful advertisement. Inquiry tests can take several forms. One way could be to place the same offer in different copies in different issues of the same magazine/newspaper. These offers are keyed to a specific advertising copy. If the number appeals more to readers. Another variant could be to give the same offer in different advertising copies that appear in different newspapers or magazines. This assumes that there are only negligible differences among different media. However, this may not be the case. Sometimes, the same offer is made through two pieces of copy. One piece of copy is carried in half the copies of the 155 newspaper or magazine and the second piece of copy is carried in the remaining half. Inquiries received are then linked to the two pieces of copy.
Inquiry tests can be developed in the form of controlled experiments to ascertain the impact of an advertisement copy. However, one has to exercise great care in isolating the effect of other factors from that of advertising.
Simulated Sales Tests: These tests expose prospective consumers to different pieces of copy through point-of-purchase displays or direct mail. Thus, one may select two groups of similar stores where two alternative pieces of copy are displayed at the entrance or at some other place in the store. Sales of the product in question are measured both before and after the display of copy in the two groups of store. The copy in those stores which have registered a higher increase in the sale of the product over time is regarded as a better copy. Likewise, comparisons can be made between two pieces of copy using direct mail.
While these tests are both more economical and simpler than actual sales tests, one major limitation is that there is no certainty that the advertisement when actually given will have the same result as at the time of the test.
Day-after Recall Tests: These tests are generally undertaken for television commercials. The test involves an on-air exposure of a commercial in a couple of cities. This is followed by a telephonic enquiry of the respondents to find out if they can recall the message. The aggregate recall score that is arrived at is compared with a standard score based on similar studies. If the score given by the commercial is higher than the standard score, it is inferred that the advertisement is useful and should be telecast on a larger television network.
The main advantage of this test is that it is performed in a natural setting. Moreover, a proper sample design can be used in this method. In contrast, the major limitation is that it turns out to be a test of the respondent’s ability to remember. This does not necessarily establish that the respondent will behave in a different way as a result of watching the commercial on television. How far can be recall be related to a change in the respondent’s attitude and behaviour? This is a pertinent question which is difficult to answer in the context of day-after recall tests.
‘After’ Tests: There are three methods that are frequently used to test an advertisement after its formal release These are recognition test, recall test and sales test. They are described below.
Recognition Tests: These tests are carried out with respect to a printed advertisement and commonly referred to as a readership study. Here, the respondents are asked if they have read a particular issue of a magazine. They are further asked as to what they saw and read. Generally, the respondent is shown a particular page of the magazine and then the following measures of recognition are taken
- Noted: the percentage of readers who have seen the advertisement earlier.
- Seen-Associated: the percentage or readers who read a part of the advertisement which indicates the brand or advertiser.
- Read Most: the percentage of readers who read a major part of the advertisement.
Scores are assigned to these three measures and overall scores are determined for all the advertisements contained in a particular issue of the magazine. These scores are then related to the expenditure incurred on the advertisement. In this way, cost ratios can be determined.
The recognition method has certain limitations. Some respondents may confuse specific advertisements with similar or identical advertisements seen elsewhere. Respondents may forget having seen an advertisement earlier or falsely claim that they have seen it.
Recall Tests: In this method respondents are asked to recall specifics of the advertisement. In the foreign countries, there are some advertising agencies that offers a post-testing readership service. To begin with, copies of test magazines are sent to a sample of respondents who are asked to read them in a normal manner. Telephone interviews are held on the following day. Respondents are read out a list of advertisements and asked to identify those they remember and the extent to which they are able to recall. Thus, scores are assigned to the ability of the respondent to remember the name of the product, the underlying message contained in the advertisement and their favourable attitude regarding the advertisement.
Recall tests, no doubt, go beyond recognition tests but it is difficult to say that recall scores indicate the desired consumer behaviour. Recall scores may be high and yet there may not be any perceptible change in the consumer behaviour with respect to the product in question.
Sales Tests: This method measures the effect of an advertisement on the sale of the product. The assumption is that changes in sales are as a result of the advertisement. However, as there are several factors influencing sales, one has to be extremely careful in establishing a relationship between advertising and sales. It is desirable to isolate the influence of other factors while determining the impact of an advertisement on the sale of the product. Experimental studies can be designed to study the impact of an advertisement on sale.
After having discussed the ‘before’ and ‘after’ tests, we now turn to two recent studies on advertising research in the Indian context.
Examples of advertising Research Studies in India: An advertising research study dealing with Santoor, a brand of toilet soap was done by Ravi Menon and Ashutosh Sinha. Simulated Test Marketing (STM) methods were used to choose between two alternative advertising themes for the relaunch of Santoor.
Santoor was launched in 1986. In early 1989, a relaunch of the brand was being contemplated. At this time its distribution was not yet fully national. The few states where Santoor had been quite successful contributed to a large part of the brand’s sales volume.
A number of changes were now contemplated: a new perfume and a new pack design, while distribution was to be extending to smaller towns and to all the states in the country. Finally, a change in the advertising was also planned: the brand was to be positioned ‘for the modern women’.
The advertising agency, Ulka, came up with two alternative appeals for advertising:
‘Romance’: to have soft fragrant skin that makes one attractive to men.
‘Young Skin;” to have youthful skin.
There alternative advertisements were developed for the ‘young’ skin’ concept, while one was developed for ‘romance’.
In the first phase of the study, qualitative research using focus group discussions was conducted in Bombay, by Probe Qualitative Research (PQR). Of the three alternative advertisements for ‘young skin’, one ad (i.e. ‘Bookshop’) was found most appearing. The lone ad for the ‘romance’ concept was also shortlisted through qualitative research. The two ads (one for each concept) thus shortlisted are described below:
Aamir Khan (Romance0: The ad features a dreaming Aamir Khan (a film star). Bookshop (young skin):
The mother of a young girl is mistaken for a college student at a bookshop. The question to be probed was: Will Aamir Khan actually succeed in generating a higher increment of trail?
An adapted form of a STM model was employed to answer this. The study was conducted in Bombay (where the focus was on the potential new tiers for the brand) and Cochin (where the main interest was the existing user base: Kerala being a strong market for Santoor). Two panels of respondents were recruited. The Aamir Khan ad was shown to one panel, while the other panel was exposed to the Bookshop ad. The exercise was carried out over the following five stages:
Stage I: Women who met the target group definition were recruited. The recruitment interview also assessed the brands of soap ever tried and those currently used. Preferences between brands were also measured using a Constant Sum Preference scale.
Stage II: Respondents were brought to a central location. They were shown advertising for Santoor as well as other ads in a clutter of commercials, after a short television programme.
Stage III: A coupon worth Rs.12 was handed over as a ‘gift’. The ladies were then led to a stall that stocked Santoor as well as several other brands. They were free to buy whatever they wished, or not buy anything gat all, in which case they could collect cash in exchange.
Stage IV: An optional in the STM procedure. In order to obtain a detailed assessment of the advertising itself, the test advertisement was screened once again, followed by a short interview. At the end of this, the brand (Santoor) was given as a gift to those who had not purchased it at the simulated shop.
Stage V: After a lapse of a week, the respondents were once again interviewed at their homes for their evaluation of the product. Their preference between brands was measured once again, this time including Santoor in every instance.
The Total Sample Size Was Over 500: It may be noted that the STM system is based on the ascertain that if we know something about a person’s attitudes, we can make a statement about his or her likely behavior. So if a person’s preference between brands is known, then we can say something about his relative chance of buying these brands. The study concludes that as the trial part of the model works well, this is a good method for advertisement testing. The trial at the simulated shop offers an absolute standard which is lacking in any kind of ‘intention to buy’ scale.
Another study done by Sankara Pillai examines the impact of clutter on advertising viewer ship and recall. One of the major problems with regard to television media planning in India is the dichotomous choice between two types of vehicles. On the one hand, there is the option of buying an extremely high viewership programme that has a string of 30-40 ads, and on the other the choice is that of a moderate viewership programme with only a moderate clutter level.
The study be Sankara Pillai uses Near-Coincidental Interviews and a series of controlled Forced Exposure Tests to provide measures of the effects of clutter on advertising. Before embarking on a study it was necessary to define how exactly clutter affects advertising. It was believed that there would be two influences at work. Firstly, the viewer may take advantage of the predictably long capsule to attend to other task. Thus, he may switch on the set later than the announced programme time or alternatively, even if the set is switched on, he may not be physically present in front of it. While the former situation implies a non-response to all of an audio There could also be intermediary situations in the quality of exposure, with the viewer being physically present in front of the set but doing something else, which results in varying degrees of attention being paid to the screen.
The second effect of clutter would be one of notice ability and registration of the advertising message. There is some evidence to show that between 1965 and 1981, a period in which the number of ads aired increased greatly, the recall levels for the last ad seen on television declined from 18 per cent to 7 per cent. Krugman hypothesizes that this could be due to perceptual defence created by the viewer in the form of increased ‘distraction opportunities’. Evidently, any study on the impact of clutter would also need to take into account the effect of clutter on notice ability and registration of the advertising message. The study as designed by the IMRB, thus, had two separate components to measure each of these two effects. The method for measuring the first effect of clutter on set switching-on behaviour consisted of Near-Coincidental Interviews. Respondents pre-selected on demographic quotas were interviewed within 30 minutes after the conclusion of the programme. This was done to avoid interrupting the viewing which could lead to non-cooperation on the respondent’s part. The interview ascertained whether the programme was watched and if it was, whether the viewing commenced from the previous programme, the advertising capsule or after the commencement of the programme. If the viewing commenced during the capsule the exact time of starting the viewership was anchored by providing the respondent with the sequential list of advertisement telecast before the programme. The same technique was adopted to monitor the movement of the viewer out of the room, if at all, after commencement of viewing. For non-viewers of the capsule the opportunity-to-hear was also determined. Data were collected for 23 programmes in this manner with a sample size ranging between 200 and 400 for each programme. The study was carried out in nine places – four large metropolitan cities and five smaller towns. Of the 23 programmes, 11 belonged to the high clutter category while the remaining 12 were of the low clutter category.
For measuring the second effect, the impact of the clutter on notice ability and recall, a series of Forced Exposure Tests were carried out. Matched panels were exposed to a test ad (being aired for the first time) either in a low clutter capsule of 12 ads or in a high clutter capsule of 30 ads. All ads were in Hindi, of 20 seconds duration each, and represented major product categories advertised on TV. In either case, the ad capsule preceded 20 minutes of a popular programme. The position of the test ad was varied within the capsule to measure the effect of position. A total of 800 people formed the sample for this component of the study, with one half of the sample being exposed to the high clutter capsule and the other half being exposed to the low clutter capsule.
The study clearly establishes that the impact of clutter is something that media planners can no longer ignore. This is especially true when planning TV media schedules for products targeted at certain segments, like the older age group, where the impact of clutter seems to be maximum.
Of the two effects of clutter on advertising, the impact on advertising notice ability and recall is much greater than that of the erosion of advertising viewership. Any study on the impact of clutter would, therefore, need to take into account both these aspects.
These examples of advertising research studies indicate that with the increasing importance of advertising in India, certain aspects of advertising are studied by professional marketing researchers. The studies also reflect the sophisticated methodology used in seeking out the right answers. It may be emphasised that these are only illustrative examples and in reality a wide variety of advertising problems are taken up for research from time to time. However, a good number of studies done by consulting firms are not made public because of the condition of confidentiality stipulated by the sponsoring organizations.
Need for More Systematic Copy Testing Research: Despite the improvement that has taken place in copy testing, there is a strong case for undertaking it in a far more systematic manner. Ostlund, Clancy and Sapra reached some interesting conclusions in this regard. One of the conclusions was that although advertises and agencies were spending a great deal of their money on copy testing services, there was hardly any assurance of the validity of this method. According to them the inaction concerning the assessment and improvement of copy-testing method performance is probably due to the lack of minimum standards for reliability, sensitivity and validity for either TV copy-testing methods or print methods.
