Agency: Bartle Bogle Hegarty Author: Nick Kendall

Hagen-Dazs: Dedicated to Pleasure

INTRODUCTION

In 1991 Häagen-Dazs became the most talked about ice cream brand in the country. Every national newspaper wrote about it and its advertising.

This paper does not talk about that fame. Rather it talks about the thinking and facts behind that success, and describes how Häagen-Dazs became clear brand leader in its sector.

In the first section, we will put Häagen-Dazs into context versus its competitors and describe how the advertising strategy and creative work were developed.

Next, looking at the effect of advertising, we describe the sales growth from October 1990 to February 1992 and attempt to isolate the effect of advertising on sales, versus those due to distribution and seasonal growth. We do this using both econometric analysis, and analysis of sales from some 'core' outlets that stocked Häagen-Dazs both before and after advertising.

Finally, we will attempt to show how we believe advertising worked; first by creating high levels of awareness and trial, and second by positioning Häagen-Dazs successfully as a sensual and sophisticated adult treat and a new gold standard in the market.

THE UK TAKE HOME ICE CREAM MARKET

The UK ice cream market is underdeveloped compared to many in the world. In the USA consumption of ice cream is three times that of the UK.

The reason for this is not a lack of sunshine (see Sweden Figure 1). The problem has been that most manufacturers' vision of the potential of the market was limited to ice cream as a child's product.

Most take home brands, therefore, are designed; to act as 'default desserts'. Quality is not a key dimension. Format is more crucial. The products come either in large 'soft scoop' tubs, or multi-packs of hand-held products, eg Cornetto (see Figure 2).

By 1991 the market had started to change and recognised the potential for better quality adult products. Vienetta offered an adult version of default desserts, and in 1990 the Mars ice cream bar raised the quality stakes in the multi-pack hand-held sector.

The market's premium sector proper (excluding complete desserts such as Vienetta, and premium multi-packs such as Mars) had also been showing signs of life for the last few years (see Figure 3).

AGB's definition of 'premium' ranges from products such as Wall's Gino Ginelli to the two brands that stood at the top of that sector and represented the premium ice cream sector proper Loseley and New England.

These represented the gold standard of the ice cream market in the late 1980s, charging around £1.99 for half a litre.

Both brands had been available for many years (Loseley for 18 and New England for 15 years). Their marketing strategies had been classic 'seed branding', involving many years building credentials in delicatessens and premium independent food shops before then moving into multiple grocers.

It was against these two brands that Häagen-Dazs was primarily competing.

HÄAGEN-DAZS' MARKET OBJECTIVE

Häagen-Dazs believed the market for premium adult ice cream was still in its infancy and that there was the opportunity to raise the quality stakes even higher ie to develop a super-premium sector.

The company mission was to create a new gold standard and become the ultimate ice cream in the market.

At £2.99 per 500ml the intention was that Häagen-Dazs should command a massive premium of exactly £1.00 versus its two closest competitors.

The basis of Häagen-Dazs' self-belief lay both in its conviction that it had a superior product to its nearest competitors and, equally importantly, that both Loseley and New England had a number of image limitations which might be exploited.

The image limitations of current premium brands

The initial quantitative research identified that the key users of current premium offerings were ABC1, 2545 year olds, both male and female, with high disposable income (eg no kids) and high terminal education age (see Figure 4).

Qualitative research amongst this group revealed that existing premium ice cream was perceived to be primarily about formal, social occasions; dinner parties, Sunday family dinners, etc.

In this restricted context a brand's role was to signal premium in a very overt way. The brand was about impressing others. Loseley, for example, was drawn by consumers as the epitome of traditional England all Barbours, Range Rovers and country halls.

New England, though branded as American, carried many of the same signals eg pack graphics showing beautiful pictures of decanted ice cream.

The target identified was uncomfortable with this imagery. They were modern and sophisticated in their use of quality food. They did not feel it should be reserved only for formal occasions and for serving to others. They were just as happy eating good ice cream sitting around in an old pullover and relaxing privately.

There was an opportunity therefore to use advertising to rewrite the rules of premium ice cream by deformalising and contemporising its imagery.

THE HÄAGEN-DAZS ADVERTISING STRATEGY

After limited tests in Sainsbury the objective in 1991 was to embark on a full scale brand launch in the south of the country London, Anglia and TVS supported with full scale advertising. This area was the most competitive in the country but accounts for 51% of the market. It is at this point we helped develop the advertising this paper will analyse.

The advertising was designed to exploit our competitors' weaknesses. Early in the process we decided not to attempt to 'outbadge' Loseley and New England.

We avoided lifestyle images of Häagen-Dazs featuring in posh flats, sitting next to champagne glasses, or being eaten in the back of Rolls Royces!

Instead, we decided to concentrate on the product's credentials and the new type of individual premium experience created by it.

As a product Häagen-Dazs was associated primarily with richness, creaminess and smooth texture and intense flavour.

It created a distinct Häagen-Dazs moment, that was described by consumers as 'languorous', 'sensual', 'dream-like'. People described a feeling of hovering and a sense of 'time stopping still'.

Häagen-Dazs became not just another quality ice cream manufacturer but a very special purveyor of a particular mood or emotion.

This Häagen-Dazs moment was most often described as a moment to be savoured and enjoyed solus either as a reward or as compensation.

But this usage also carried some negative connotations of 'comfort eating' and guilty 'pigging out'.

However, in research there were hints of a different, more attractive, potential benefit; that of people sharing Häagen-Dazs and the Häagen-Dazs experience. Consumers talked of sharing a spoon, feeding each other, and of 'mellowing out' together in front of their favourite video with the lights off and curtains closed... and at this point people began to giggle knowingly.

This area offered Häagen-Dazs the chance to create a unique advertising language for ice cream. Solus reward eating was the cliché of premium food and would not have been as brave and as different as the product itself. The idea that Häagen-Dazs created a mood of sensual intimacy between adults was a far more interesting area for the brand to carve out and make its own.

The creative brief

Why are we advertising?
To position Häagen-Dazs as the new gold standard in the market by referring to the immense pleasure of the Häagen-Dazs experience.

Who are we talking to?
ABC1, 25 plus, single, male or female looking for new interesting quality tastes. They demand the best but are not precious or pompous about it.

What must the advertising say?
Häagen-Dazs the ultimate in intimate pleasure.

Why should the consumer believe it?

What Häagen-Dazs put into the pot:
dedication to perfection
extraordinary ingredients

What people take out of the pot:
an extraordinary personal pleasure

What tone of voice?
Adult, sensuous

The creative idea

The advertising idea flowed naturally out of the research.

We decided to juxtapose what Häagen-Dazs put in the pot (as evidenced by actual quotes taken from their company literature) with what consumers got out of the pot.

The approach was summed up in the line 'Dedicated to Pleasure' which was a tongue in cheek rewrite of the original company slogan 'Dedicated to Perfection'. In total four executions were developed.

One of these, 'Feel It', we later recognised carried too many connotations of eating alone, and therefore was not used in the second phase.

The media strategy

87.5% of ice cream advertising in 1991 was on TV. TV was a budget option.

However, we decided to break away from this traditional means of media support, despite many attractive offers from TV companies for Häagen-Dazs as a new product.

Rather we chose to view media as part of the process of positioning Häagen-Dazs as a 'mood purveyor'.

First, we knew we wanted to achieve a feel with this advertising that could itself be savoured and enjoyed at leisure, just like the product.

Second, we believed the intimacy of the experience could be better illustrated through personal communication. Television is often a family or social medium. We wanted the communication to be private and not expose it to the comment and reaction of third parties.

Third, in keeping with the desired brand personality, we wanted to achieve a relationship with our target audience gradually, rather than blast them with a one-off message.

The solution we chose was to spend the bulk of our monies against the weekend colour supplements and in particular the weekend review sections of the up-market press.

Though we chose national titles these quality papers had a regional bias in terms of coverage. This coincided with the bias in our distribution and sales effort for the launch.

TABLE 1: TARGET MARKET COVERAGE OF PROPOSED SCHEDULE BY REGION

Index of
coverage
London118
TVS110
Anglia101
National coverage100
Source: NRS

The coverage achieved in the rest of the country was seen as investment for the (very near) future.

TABLE 2: HÄAGEN-DAZS SUMMER SCHEDULE

Week commencingJulyJulyJulyAugAugAugAugSept
15222951219262
Review section
IOS Review
Times Review
Supplements
The Observer
The Independent
The Sunday Times
The Telegraph
You
Evening Standard
Women's magazines
Vogue(Sept issue)
Tatler(Sept issue)
Elle(Sept issue)
Marie Clare(Sept issue)
Vanity Fair(Sept issue)
Cosmopolitan(Sept issue)
Source: BBH

This first burst started w/c 15 July and finished by w/c 2 September. It achieved 67.4% coverage of ABC1, 2545 year olds with an average frequency of 4.8 OTS.

TABLE 3: HÄAGEN-DAZS CHRISTMAS SCHEDULE

Week commencingDec 7Dec 14Dec 21
IOS Review
Saturday Times Review
The Sunday Times
The Guardian Wk
The Independent Wk
Source: BBH

After the initial first burst we returned in December in anticipation of the Christmas season.

This second burst achieved 38.1% coverage with a frequency of just 2.5 OTS.

The MEAL estimate of the spend was £900,000, which gave Häagen-Dazs only a 6.0% share of voice across the whole year.

MEAL estimates are greatly exaggerated and, in fact, Häagen-Dazs' actual spend was roughly a half of this figure.

THE GROWTH IN HÄAGEN-DAZS' SALES IN 1991

In 1991 sales of Häagen-Dazs rose by 398% versus 1990.

Figure 5 shows Häagen-Dazs' overall retail sales, across the whole country, measured in terms of indexed ex-factory sales.

Figure 6 shows Häagen-Dazs' growth in multiple grocers in the launch area only using Nielsen Scantrack. We will use these data to examine the effect of advertising in more detail. It is a robust source based on EPOS data and therefore acts as a good proxy for the whole market.

The Nielsen sales show a steeper incline than the sales overall. The multiples represented a new area for the brand and therefore sales grew from a smaller base. The result is that Häagen-Dazs multiple grocers in the area at the beginning of the launch accounted for less than 20% of Häagen-Dazs' overall retail sales. They grew to represent around 50% in the final periods.

Figure 7 shows this growth in the context of what was happening in the total market in multiple grocers in the launch area. Dairy market is Nielsen's equivalent to AGB's 'premium' ice cream sector and includes the same range of brands.

The tables show how Häagen-Dazs grew across the year in a number of distinct phases.

First, the brand sales gradually increased in the early part of the year.

This growth accelerates in the periods immediately before advertising, as distribution in anticipation of advertising is added, particularly by the multiples.

The first burst of advertising accelerates growth again.

Sales then fall as the market declines due to seasonality, though this happens at a much slower rate than the market overall. Thus brand share continues to rise.

The second burst of advertising in December increases sales growth again.

ISOLATING THE ADVERTISING EFFECT

We do not intend to claim that advertising created all the growth shown above.

Growth was, in fact, the result of three key variables: market seasonality, availability of product ie distribution, and finally, but by no means least, straight increase in consumer demand due to advertising effect.

Figures 8 and 9 show how the first two of these variables were also increasing during the advertising period.

It is clear that Häagen-Dazs benefited from seasonal and distribution increases.

The effect of distribution and seasonality could be isolated by looking at Häagen-Dazs' rate of sale and Häagen-Dazs' brand share separately. However, both these measures look at the effect of the different variables in isolation from each other. Rate of sale isolates the effect of distribution but not seasonality, and vice versa.

The simultaneous growth in the different elements suggests a far more complex interplay of variables. We therefore used econometric analysis to help disentangle them.

THE ECONOMETRIC ANALYSIS OF HÄAGEN-DAZS' GROWTH

It was first necessary to identify what we believed to be the underlying process that led to the brand's success. This theory was then tested against the data, and quantification of the relationships as derived using econometric methods. The diagram below attempts to show the hypothesis in a simplified pictorial form (see Figure 10).

This theoretical model suggests that advertising had a number of direct as well as indirect effects on sales.

First, advertising for Häagen-Dazs, as with many new products, was part of the special trade negotiations which helped the brand secure the extra distribution before advertising even started (this distribution growth was seen in Figure 9).

Though it is tempting to claim advertising played some part in the securing of this distribution, we do not. The econometric analysis discounts any claim for sales due to direct trade negotiations by including these in the equation as a separate variable.

Second, advertising created direct increased consumer demand for Häagen-Dazs. The econometric analysis posits this as the first order effect of advertising.

Third, advertising's effect on consumer demand and the fame it created for the brand helped encourage trade buyers either to increase distribution, or provide distribution where before they had refused. In other words for every 1% of extra sales created by advertising x% of extra distribution was created. In effect this is a 'success breeds success' explanation of what happened, which has some appeal to common sense.

Häagen-Dazs' Director of National Accounts confirmed this change of attitude.

Before advertising broke the trade were obviously wary. They were naturally suspect of our claims that we could create a new market for a different type of super-premium brand. After advertising doors opened that had been closed. Buyers started calling us. As a result test stores became full listings, and stores who had initially declined Häagen-Dazs changed their position.
Mel Bugler, Director of National Accounts

Two equations are suggested by the theory one that explains advertising's role in volume sales, and one its role in distribution. The actual data fitted this theory well. Approximately 99% of the variation in both sales and distribution data is explained by the models (see Figure 11).

In effect these two equations 'interact' creating a ripple effect. Advertising first creates extra sales, extra sales then create extra distribution (see Figure 12). Advertising continues to create sales, but these are added to by the increased sales due to distribution. The result is yet more increased distribution and so on, and so on.

The effect of the advertising is shown below. Figure 13 shows sales due to direct consumer demand created by advertising. Figure 14 shows sales resulting from distribution growth due to advertising. These are estimated by looking at actual sales versus simulations of sales without advertising.

THE TOTAL VALUE OF SALES DUE TO ADVERTISING

Advertising created a 59.7% increase in volume over the eight month period since the first burst of advertising. 78% of this increase was due to the first order effect of advertising on consumer demand.

This is equivalent to a projected increase in value sales, as measured by Nielsen, across the full year of roughly £440,000.

This is a valid calculation as the model indicates that there is no evidence that the sales effect due to advertising has decayed. It is as if the demand for Häagen-Dazs has been shifted to a once and for all higher level.

The figures above relate to sales through launch area multiples only. Examination of the 'core' outlets (see next section) suggest sales increased across Häagen-Dazs' total distribution base. Distribution also seemed to grow across the country in roughly the same proportion (from around 3% to 32% on multiples across the country for example according to Nielsen). It seems fair to assume therefore that advertising achieved the same levels of sales increase across the country.

We have calculated earlier that multiple grocers accounted for just 50% of Häagen-Dazs' total retail sales. The sales effect due to advertising can therefore be roughly doubled.

This gives a figure which shows a sales uplift due to advertising within the first year of nearly two times actual negotiated media spend.

TESTING THE MODEL

In addition to the above evidence we have a totally different data source that externally validates the modelling analysis, and provides further supporting evidence as to the contribution of the advertising. This data source is a set of 'core' stores (see Figure 15).

These core stores are those of a large deli/grocer group based in London. This group has over 20 stores and Häagen-Dazs was stocked in all of them from October 1990. Sales data were collected on a monthly basis.

This sudden rise in July cannot be attributed in any way to distribution growth. Seasonality played some part. However, it also represents some kind of once and for all leap in consumer demand. This is seen even more clearly if we look at year on year sales across the same monthly periods ie October on October, November on November, etc. This discounts any seasonality effect.

TABLE 4: % INCREASE IN £ SALES YEAR ON YEAR

+ %
Oct51
Nov18
Dec44
Jan41
Feb51
Source: Accounts for London based deli/grocer group

Over the five months this represents an average 41% increase in rate of sale. This increase is roughly equivalent to that indicated by the econometric analysis of increased consumer demand excluding distribution effect.

DISCOUNTING OTHER FACTORS

The econometric analysis has discounted seasonality and distribution. This section will consider some of the other potential creators of consumer demand.

The economy

The analysis did not include figures for the general economic climate. However, common sense tells us that the climate did not exactly work in the brand's favour.

The launch took place in the middle of the longest recession since the Second World War. The result was a decline in consumer spending which, if anything, might be expected to have dampened the effect of consumer demand (see Figure 16).

Market growth

The econometric analysis works on share and therefore discounts general market growth effect. However, we should consider whether Loseley and New England in particular also increased share ahead of the market. In fact these two brands lost share over the year (see Figure 7).

There was no growth at the top end of the market that benefited Häagen-Dazs.

COMPETITIVE ACTIVITY

Advertising and promotion

General competitive activity increased dramatically in the years up to 1991.

TABLE 5: ADVERTISING SPEND 1987 TO 1991

£000
19875,673
19887,275
19898,847
199016,814
199114,471
Source: MEAL

In terms of immediate competition, Loseley devoted its marketing activity to new packaging and POS. New England actually advertised for the first time in TVS, using TV. They achieved 837 TVRs: 68% cover at 4+ OTS (Source: DDS).

We can see from Figure 7 that the brand achieved no overall increase in brand share.

This is clear evidence that it is not just a case of any advertising achieving success. The advertising has to be right.

Distribution

Loseley and New England maintained the same levels of sterling distribution throughout the main part of the year (New England declined slightly in the last three periods). See Figure 17.

Therefore it cannot be argued that Häagen-Dazs benefited from the distribution decline of its nearest rivals. Häagen-Dazs won in terms of straight consumer demand. Häagen-Dazs' distribution lagged behind that of Loseley/New England making its brand leadership even more surprising.

Price

Häagen-Dazs' marked premium was maintained throughout the periods discussed. No price cutting or promotional discounting was employed to kick start the launch, nor at any point after (see Figure 18).

Sampling and shops

Through the period Häagen-Dazs opened further shops in Oxford, Bath and Heathrow. These acted as sampling points. Häagen-Dazs also carried out sampling to support the launch.

In order to measure the contribution of these elements we used tracking to monitor their effect. This took place before advertising, and then after each burst.

The proportion of the people receiving free samples remained roughly the same, or even dipped during the period of advertising.

TABLE 6: THE CONTRIBUTION OF SAMPLING (%)

PrePost 1Post 2
TotalAwareTotalAwareTotalAware
(300)of Häagen-(300)of Häagen-(300)of Häagen-
Dazs (32)Dazs (75)Dazs (77)
Receipt of free
sample in
supermarket-415-1
Receipt of free
sample
elsewhere11015310
Base: 300 ABC1 adults, 2545 years, London/Anglia/Southern
Source: Consumer Focus Tracking Study
Note: ( ) = Sample size

We also asked respondents how they had first become aware of Häagen-Dazs.

TABLE 7: HOW FIRST BECOME AWARE OF HÄAGEN-DAZS (%)

PrePost 1Post 2
(32)(75)(77)
Saw product in
Häagen-Dazs shop313
Base: All aware of Häagen-Dazs
Source: Consumer Focus Tracking Study
Note: ( ) = Sample size

The proportion of people who had come to the brand through the shops declined during the advertising.

Our conclusion is that the massive rise in demand was created above and beyond the contribution of sampling and shops.

THE EFFECT OF ADVERTISING ON HÄAGEN-DAZS AWARENESS

A tracking study was carried out amongst our core target market in London, Anglia and the South in three waves:

Pre pre-advertising (June 1991)
Post 1 post-1st burst (Sept 1991)
Post 2 post-2nd burst (Jan 1992)

This allowed us to look at rises in advertising awareness, awareness, usage, image and propensity to trial.

Advertising awareness

Rises in advertising awareness were very strong (remembering the size of the budget), especially amongst those directly exposed to the advertising message (ie readers of papers/magazines on the actual schedule).

TABLE 8: AWARENESS OF HÄAGEN-DAZS ADVERTISING (%)

PrePost 1Post 2
Non-Non-Non-
SchedulescheduleSchedulescheduleScheduleschedule
TotalreadersreadersTotalreadersreadersTotalreadersreaders
(300)(193)(107)(300)(179)(121)(300)(99)(202)
Spontaneous
awareness00071034111
Prompted by
name1111419610234
Prompted by
pictures22224321222384
Base: 300 ABC1 adults, 25-45 years, London/Anglia/Southern
Source: Consumer Focus Tracking Study
Note: ( ) = Sample size. Statistical error: 4-5% (see appendix)

The total figure declines due to the reduced coverage of the advertising in second burst (see media schedules) which was half that of the first. Awareness therefore holds amongst those potentially exposed to the campaign, but declines amongst those not on the schedule.

This effect is exaggerated as the market enters the Christmas season.

These two factors deflate the 'total' figures in the final wave of tracking.

It is encouraging, however, to see that awareness of those who had chance to see the second burst of advertising (ie 'readers on schedule') stays high, or even increases.

Brand awareness

The growth in advertising awareness was mirrored by growth in spontaneous and prompted awareness, again especially amongst schedule readers.

TABLE 9: AWARENESS OF HÄAGEN-DAZS PRE AND POST ADVERTISING (%)

PrePost 1Post 2
Non-Non-Non-
SchedulescheduleSchedulescheduleScheduleschedule
TotalreadersreadersTotalreadersreadersTotalreadersreaders
(300)(193)(107)(300)(179)(121)(300)(99)(202)
Spontaneous
awareness220111475122
Prompted
awareness
(name only)11128253313254317
Base: 300 ABC1 adults, 2545 years, London/Anglia/Southern
Source: Consumer Focus Tracking Study
Note: ( ) = Sample size

Trial and intention to purchase

The growth in awareness for Häagen-Dazs was mirrored by increases in trial, with higher figures amongst readers on the schedule actually exposed to advertising.

TABLE 10: CLAIMED USAGE OF HÄAGEN-DAZS PRE AND POST ADVERTISING (%)

PrePost 1Post 2
Non-Non-Non-
SchedulescheduleSchedulescheduleScheduleschedule
TotalreadersreadersTotalreadersreadersTotalreadersreaders
(300)(193)(107)(300)(179)(121)(300)(99)(202)
Eaten last
12 months666111567125
Eaten last
3 months2148113554
Base: 300 ABC1 adults, 2545 years, London/Anglia/Southern
Source: Consumer Focus Tracking Study
Note: ( ) = Sample size

Claimed intention to purchase also increased amongst those aware of Häagen-Dazs (see Figure 19).

HOW THE ADVERTISING POSITIONED HÄAGEN-DAZS

The advertising also helped carve out a clear point of view for the brand.

First, the advertising positioned the product as the gold standard. In the tracking the consumers were asked whether they agreed/disagreed with a number of statements, such as 'best quality ice cream available'.

Häagen-Dazs achieved pole position versus its two closest competitors, after just 3 years in the market and six months of advertising.

TABLE 11: % AWARE OF EACH BRAND AGREEING THAT PRODUCT
IS 'THE BEST QUALITY AVAILABLE'

Häagen-Dazs51
Loseley44
New England32
Base: All aware of brands
Source: Consumer Focus Tracking Study
Note: Sample size HD = 75, L = 88, NE = 57

The objective of the advertising however was not just to ensure we became the new gold standard in rational terms. It was also to create a whole new set of image associations.

Qualitative research showed that we clearly achieved our objectives of creating a new language for Häagen-Dazs.

The campaign has had considerable effect on perceptions of Häagen-Dazs' image. It clearly positions Häagen-Dazs as premium quality, special, adult, modern, worthy of attention, exclusive and a whole new different angle on ice cream.
SRG Research, Oct 1991, Base: ABC1 adults, 2545 years

The list of words below shows the initial reactions to the advertising in the groups. Respondents were asked to write down on paper the first words that came into their head.

TABLE 12: INITIAL RESPONSES TO ADVERTISING COMMUNICATION

SensualWickedFunNaughty
EroticIrresponsibleSmoothSensation
SexualNaturalYouthfulHappy
MovingSexySeductiveIntimate
PleasureDesirablePassionate
SharedTogethernessRaunchy
CoolPrivateStunning
VibrantSophisticatedIndulgent
Source: SRG Research, Oct 1991
Base: ABC1 adults, 2545 yrs

The advertising succeeded therefore in positioning Häagen-Dazs as a sensual, sophisticated adult treat.

HOW ADVERTISING MADE HÄAGEN-DAZS FAMOUS

Before advertising there was little press coverage of Häagen-Dazs. Most was trade based except for two or three paid-for consumer advertorials in June.

After advertising national consumer press coverage soared (see Figure 20).

A content analysis of consumer articles reveals that, of the 263 consumer articles, 42% mentioned the advertising. Another 14% of the articles actually featured one or more of the advertisements, as well as then going on to discuss the brand and advertising. Extra PR gave 52% coverage, at 2.2 OTS (analysis by Biss Lancaster).

Press coverage continued well after the campaign finished. The latest piece of 'free' advertising was the National Press Association's use of the case history in their campaign on the benefits of press. This took place in March 1992. It is estimated that £1million at rate card was spent on this one execution. Its effect is, however, outside the periods covered by this paper.

CONCLUSION

This paper, we believe, has demonstrated how advertising helped Häagen-Dazs gain brand leadership of the take home premium ice cream market.

First, advertising with only a 6.0% share of voice, made Häagen-Dazs the most talked about ice cream brand of the year, indeed the 'New Product of the Year' according to The Marketing Society.

The advertising first created consumer excitement, then PR, then trade excitement, and consequently more and more consumer interest. The whole launch gained an unstoppable momentum.

On top of that, advertising helped create an absolutely clear brand point of view for Häagen-Dazs. Everyone who knows Häagen-Dazs knows what the brand stands for. Häagen-Dazs is 'Dedicated to Pleasure'. This was a completely new and unique point of view on the market, which redefined what 'premium' meant to consumers.

These brand values, both client and agency would argue, are Häagen-Dazs' future competitive advantage as new products attempt to follow. Certainly they help justify Häagen-Dazs' considerable premium versus rivals.

When we developed our advertising proposals we knew we were creating a new advertising language for ice cream in this country. We knew it broke many of the rules. Häagen-Dazs was willing to take that step and proved that they were as dedicated to producing good advertising as they were to producing great product.

We hope to have proved in this paper that they got the advertising they deserved.

TECHNICAL APPENDIX

The theoretical sales growth schema in the main text generates a number of testable predictions ie the dynamic structure, the form of the equations and bounds parameters of the variables.

Reliable Anglia/London/Southern Nielsen Scantrack four weekly data for Häagen-Dazs volume sales, distribution and total ice cream market data were available for the period October 1990 to February 1992.

Häagen-Dazs press advertising data were represented as +1 cover, for bursts 1 and 2 in the volume model, and as 0,1 variable in the distribution model. Competitive advertising was represented by TVRs.

The Häagen-Dazs advertising effect, over the available data did not decay, and entered the empirical model as zero decay adstock.

Distribution was increased by trade negotiations. The influence was captured by a 0,1 variable. The following mnemonics have been used to represent the variables.

L = logarithmic base e transformed. V = Häagen-Dazs volume. D = Häagen-Dazs sterling distribution. COMTVR = competitive advertising. COV1 = Häagen-Dazs first burst advertising. COV2 = second burst. HDMED = 0,1 variable for Häagen-Dazs advertising. TVOL = Total market volume. NEGOT = Trade negotiations.

The system can be shown to be block recursive and therefore if the Gauss-Markov conditions hold OLS will produce the BLU Estimates.

Sales volume model

VariableParametert-ratio
Constant -6.7489.2
LD(t)0.7308.4
COV1(t-1)+ 0.5*COV2(t)0.005285.1
COMPTVR(t-1)-0.000583.4

R² = 0.994, R² bar = 0.992, Regression standard error = 0.051

DW = 2.5 this value is in the indeterminate region therefore Box-Pierce Q-statistics have been provided as tests against first and higher order autocorrelation. Short Chow parameter stability and constant variance = 0.393, f(4,6). Chi-squared Q statistics, Q(1) = 1.335, Q(2) = 4.854, Q(3) = 5.997, Q(4) = 6.131 cv 9.49, Q(5) = 6.624 cv 11.1, Q(6) = 6.700.

Distribution model

VariableParametert-ratio
Constant2.05317.9
LV(t-1) - LV(t-2)0.5695.8
LV(t-2)0.3776.9
HMED(t-2)0.2424.9
NEGOT0.3996.9

R² = 0.991, R² bar= 0.988, Regression standard error = 0.052. DW = 1.987. Short Chow parameter stability and constant variance = 0.207, f(4,6). Chi-squared statistics, Q(1) = 0.002, Q(2) = 0.122, Q(3) = 2.911, Q(4) = 2.914, Q(5) = 3.013, Q(6) = 3.266.

The models fit the data and theory very well. The SE of the models are consistent with the data sampling error. All parameters are statistically significant at the 99% level and some at the 99.9% level, and stable on the basis of Chow tests. The Gauss-Markov conditions are upheld by the diagnostics.

NOTES & EXHIBITS


Haagen-Dazs: Dedicated to Pleasure

FIGURE 1: ICE CREAM AROUND THE WORLD

Source: Trade Estimates
Note: These figures are underestimated by half due to the fact that much of these countries' ice cream is artisan produced and locally consumed so very difficult to measure.

Haagen-Dazs: Dedicated to Pleasure

FIGURE 2: SECTOR SHARE BY VALUE OF THE UK TAKE HOME ICE CREAM MARKET 1990

Source: AGB

Haagen-Dazs: Dedicated to Pleasure

FIGURE 3: THE GROWTH OF THE UK PREMIUM ICE CREAM MARKET

Source: AGB

Haagen-Dazs: Dedicated to Pleasure

FIGURE 4: INDEX OF USERSHIP OF PREMIUM ICE CREAM IN LAST 12 MONTHS

Source: BBH Qualitative Research May 1990

Haagen-Dazs: Dedicated to Pleasure

FIGURE 5: THE OVERALL RETAIL GROWTH OF H�AGEN-DAZS

Source: Ha�gen-Dazs accounts, revised to create equal monthly accounting periods

Haagen-Dazs: Dedicated to Pleasure

FIGURE 6: THE GROWTH OF H�AGEN-DAZS IN MULTIPLE GROCERS IN LAUNCH AREA*

Source: Nielsen Scantrack � London/Anglia/Southern
Note: * Nielsen Scantrack splits GB into three areas: London/Anglia/Southern, WWW/Midlands/Lancashire, Yorkshire/Tyne Tees/Scotland. Data cannot be broken into individual regions due to sample size.

Haagen-Dazs: Dedicated to Pleasure

FIGURE 7: THE GROWTH IN H�AGEN-DAZS SHARE VERSUS KEY COMPETITORS

Source: Nielsen Scantrack � London/Anglia/Southern
Base: Dairy market, London/Anglia/Southern

Haagen-Dazs: Dedicated to Pleasure

FIGURE 8: MARKET GROWTH IN H�AGEN-DAZS LAUNCH AREA

Source: Nielsen Scantrack � London/Anglia/Southern

Haagen-Dazs: Dedicated to Pleasure

FIGURE 9: H�AGEN-DAZS STERLING DISTRIBUTION IN LAUNCH AREA*

Source: Nielsen Scantrack, London/Anglia/Southern

Haagen-Dazs: Dedicated to Pleasure

FIGURE 10: HOW H�AGEN-DAZS ADVERTISING WORKED


Haagen-Dazs: Dedicated to Pleasure

FIGURE 11: H�AGEN-DAZS SALES: ACTUAL VERSUS MODEL

Source: Econometric analysis

Haagen-Dazs: Dedicated to Pleasure

FIGURE 12: H�AGEN-DAZS DISTRIBUTION: ACTUAL VERSUS MODEL

Source: Econometric analysis

Haagen-Dazs: Dedicated to Pleasure

FIGURE 13: H�AGEN-DAZS SALES DUE TO INCREASED CONSUMER DEMAND

Source: Econometric analysis

Haagen-Dazs: Dedicated to Pleasure

FIGURE 14: H�AGEN-DAZS SALES DUE TO INCREASED DISTRIBUTION

Source: Econometric analysis

Haagen-Dazs: Dedicated to Pleasure

FIGURE 15: SALES VALUE GROWTH IN CORE OUTLETS

Source: Accounts for London-based deli/grocer group

Haagen-Dazs: Dedicated to Pleasure

FIGURE 16: THE UK ECONOMY 1990/91, % CHANGE YEAR ON YEAR

Source: Government statistics 1985 prices

Haagen-Dazs: Dedicated to Pleasure

FIGURE 17: STERLING DISTRIBUTION OF LOSELEY/NEW ENGLAND IN LAUNCH AREA

Source: Nielsen Scantrack
Base: London/Anglia/Southern

Haagen-Dazs: Dedicated to Pleasure

FIGURE 18: PRICE DIFFERENTIALS OF KEY PREMIUM BRANDS 1990/91

Source: Nielsen Scantrack

Haagen-Dazs: Dedicated to Pleasure

FIGURE 19: PROPENSITY TO PURCHASE HA�GEN-DAZS PRE AND POST ADVERTISING

Source: Consumer Focus Tracking Study
Base: All aware of Ha�gen-Dazs

Haagen-Dazs: Dedicated to Pleasure

FIGURE 20: NUMBER OF HA�GEN-DAZS ARTICLES PRE AND POST ADVERTISING

Source: Biss Lancaster analysis
Note: First advertisement appeared 15 July