Agency: BMP DDB Needham Author: Sarah Carter

Barclaycard: How a Bungling Secret Agent Did More Than You'd Credit

BACKGROUND

Unfortunately, rapid market growth does not continue indefinitely. When growth stops, brands need to adapt to survive.

By the end of the 1980s, Barclaycard faced a problem. The credit card market was maturing, with growth rates rapidly declining, at the same time as Britain slid into one of its longest recessions. Profitability dramatically slumped.

Barclaycard needed to reposition itself. It did so by improving the product, providing the platform to relaunch Barclaycard, and then aggressively communicating this change with a new advertising campaign and increased levels of investment. Profitability was restored.

We are familiar with advertising 'adding value' to parity products. This paper shows the advertising contribution to a business through 'adding value' to a differentiated product.

Details of Barclaycard's new product improvements had been communicated by letter to cardholders, but it took advertising to convert this knowledge into the changes in attitudes and behaviour which turned round the brand's performance.

There are few brand success stories in a recession. This is a striking example of an exception and one in a market inextricably linked to total consumer spending where it would therefore perhaps least be expected.

Barclaycard responded to exceptionally tough market conditions by investing in both the product and advertising. This can be contrasted with the strategy of the main competitor, Access, who reduced advertising support on an unchanged product.

The brave Barclaycard strategy paid handsome dividends propelling Barclaycard to brand leader for the first time in over 13 years, and restoring profitability.

A growth market 19661989

In 1966, Barclays Central Retail Services Division (CRSD, part of Barclays Bank) launched Barclaycard the first means of paying 'by plastic' with a rolling credit facility.

American Express (Amex) launched in 1962, but had no credit facility bills had to be settled each month. Amex was also targeted discretely at a small group of high-spending travellers and was not therefore a direct competitor to credit cards.

The credit card market belonged solely to Barclaycard until 1973, when a joint venture by the other major clearing banks formed the Access brand. This quickly grew, by virtue of the size of the banks' combined customer base, to become the largest brand in the market.

The market then entered a period of continuous, rapid and profitable growth. This was helped by the development of the international Visa and Mastercard organisations which enabled usage of UK credit cards abroad (Barclaycard belonged to the Visa network and Access to Mastercard). A virtuous circle was established: more people owning credit cards, and more places to use them both abroad and in the UK (see Figure 1).

Barclaycard's market share had fallen steadily for many years (as a result of other banks and building societies entering the market).

TABLE 1: BARCLAYCARD SHARE OF TURNOVER (%)

19751980198319861989
5044403734
Source: CLSB/BBA

However, Barclaycard's growth in card numbers and turnover, reflecting that of the market, had been rapid. Profits were high.

A maturing market

By 1989, a number of factors came together to cause a fundamental change in the market and as a result a total turnaround in Barclaycard's profitability.

In order to understand the impact of these changes it is necessary first to explain how credit card issuers generate income and hence profit. At the time there were only two sources:

  1. Turnover-based income: a proportion of each credit card transaction value received by the credit card issuer.

    The more cardholders spend on their credit cards, the more the issuer receives.

  2. Interest income: interest paid by cardholders on the proportion of the monthly balance not paid off each month.

    The higher this proportion, the more the issuer receives in interest.

It is the interest-paying cardholders who are the most profitable to any credit card issuer.

By the late 1980s Barclaycard faced a maturing market and on top of this, a recession. Both turnover and interest income were hit.

Turnover income

The long-term growth in credit card turnover ground to a halt as the economy entered recession. Debit cards, introduced in 1987, had also started to take off as a payment mechanism and therefore invade credit card territory. Barclaycard was still losing share, but now in a market that had stopped growing.

TABLE 2: YEAR ON YEAR % CHANGE IN REAL MARKET TURNOVER

Dec 1987Dec 1988Dec 1989
+22%+8%+0%
Source: BBA deflated by RPI

Interest income

The way credit cards were used was changing too. People were becoming increasingly wary of credit due to the recession and more sophisticated in their use of credit cards. As a result, the profitable proportion of cardholders who regularly paid interest was declining.

TABLE 3: BARCLAYCARD CARDHOLDER PROFILE (%)

19851989
Full payers1734
Interest payers4637
Dormants (non-users)1914
Others (occasional users, etc)1815
Source: Barclaycard

The effect of these changes on Barclaycard was dramatic. In 1989, Barclays Central Retail Services Division (CRSD) saw profits fall by more than a half compared with 1988, and in 1990, they made a loss.

1989 A NEW BUSINESS STRATEGY

Barclaycard had a problem. The market had changed, probably irreversibly. Barclaycard was losing share and money. It was decided to consider the long-term strategic options.

Clearly, in a static market, Barclaycard needed to find a way to turn round and increase their share of turnover if they were to grow their business.

There were two possible strategies: either to be cheaper than other credit cards or to be 'better'.

The first option was dismissed. Since all credit cards at this time were free, the only way to become cheaper was by cutting interest rates. Cardholders however are remarkably unaware of comparative credit card interest rates and as we have seen, only a third of Barclaycard holders regularly paid interest, so this benefit was not widely relevant.

Barclaycard decided to look at developing a 'better' product.

The decision to change the product

By 1989, research revealed a market of parity products. Credit cardholders saw their cards as similar if not identical and, as a result, had little or no loyalty to any particular card. They were interchangeable.

You just get out whichever card is at the top of your wallet
One's green, one's blue, after that what's the difference?
Source: TQC 1989

To grow share in a parity market, clearly the product needed to be different.

Barclaycard Marketing decided to be brave. They would pursue a differentiation strategy by upgrading the Barclaycard product. They would also go one step further and charge an annual fee to provide an additional income stream.

Barclaycard would reposition itself as the superior mass-market credit card - a platform from which to begin a new aggressive share building marketing strategy.

Marketing then set about agreeing the form of the new product: both the level of the fee and the new product features.

Extensive research among Barclaycard holders was undertaken in January 1990.

The new product

It was obvious from the start that imposing a fee would inevitably lead to the loss of some cardholders.

In research, some people reacted overwhelmingly negatively to the idea of a fee, with or without the suggested changes to the product. Although rationally the fee levels being explored were low, emotionally these cardholders were outraged at having to pay for something that had always been free.

These people tended to be the more sophisticated credit card users ('full payers') and the 'dormants' (non-users) both were not profitable. The likely response of these cardholders to a fee was to switch to a free card or even give up credit cards altogether.

Other cardholders (most were interest payers) reacted more positively. In research they could come to a more reasoned view. On balance, the benefits of Barclaycard with the additional new features outweighed what really was a minimal cost.

Providing the new product was seen as offering value to enough cardholders in total to still benefit from economies of scale, the fee provided the opportunity to create a more profitable cardholder base.

Following some quantified validation, it was concluded that the optimum product mix was:

  • an annual fee of £8.
  • a modest reduction in interest rate from 29.8% to 27.8% to maintain a fairer balance between full payers' and interest payers' income to Barclaycard.
  • free 'Purchase Protection' ie 100 days insurance against loss, damage or theft on any item bought on Barclaycard.
  • a free International Rescue Service offering help abroad including emergency cash and Barclaycard delivery if the card was lost.
  • opportunity for Barclaycard Visa holders to apply for a free Barclaycard Mastercard. This card was linked to the Mastercard (as opposed to Visa) international network. It had a separate bill and credit limit and therefore helped to organise spending more effectively.

By April 1990, a new, differentiated product was in place from which to relaunch Barclaycard. The marketing objective and strategy were set as follows:

Marketing objectives

Overall, to restore profitability for Barclaycard by:

Restructuring the cardholder base

The objective was to lose less than 20% of cardholders following fee introduction and to improve the quality of the cardholder base by restricting losses to less profitable cardholders ie: dormants and full payers.

Increasing Turnover

Barclaycard's share of credit card turnover had been in long-term decline (see Table 1).

If the decline continued at its current rate, in addition to the inevitable loss in turnover from departing full paying cardholders, Barclaycard were predicting a turnover share of around 31% in 1991 and about 30% by the end of 1992.

The objective against turnover that they set themselves was not only to halt this decline but to reverse it. Specifically, they were aiming for turnover share to be about 33% and rising by the end of 1992, rather than 30% and falling.

(Whilst 3% difference in share may not sound a lot, it is worth noting that 1% of turnover share was equivalent to about £280 million in 1990.)

Increasing new cardholders

Barclaycard wanted to halt their declining share of new cardholders and increase this share significantly, from 10% in 1989 to 15% by 1992 a difficult objective, but vital to ensure Barclaycard's long-term success.

The first two objectives (minimising losses and increasing turnover) were linked. Research had shown that the single most common reason given for keeping any credit card was the fact that it was used regularly. In addition, at a time when people were likely to be rationalising their card holding due to fee introductions, any card retained was more likely to be used.

The new Barclaycard

The new product and fee were introduced and explained by letter to cardholders in May 1990, and were promoted in leaflets included with statements over the following months.

A NEW ADVERTISING CAMPAIGN

Having changed the product, Barclaycard conducted an agency review. The brief was to develop new advertising to support the repositioning of Barclaycard. BMP were appointed in May 1990 as Barclaycard's new agency.

The role for advertising

Against the three marketing objectives the role for advertising was felt to be as follows:

Restructuring the cardholder base

In the case of 'Die-hard' fee-averse customers most of these were full payers or dormants there was little advertising could do to stop them leaving. (In any case, they were unprofitable to Barclaycard.)

In the case of non 'Die-hard' fee-averse customers most of these were interest payers and hence profitable. We believed advertising could play a role here by convincing them of the increased value of a Barclaycard hence cushioning the blow of the £8 fee and preventing them from leaving. (Credit balances can be transferred between cards. They are not a barrier to moving and hence interest payers could give back their Barclaycards if not persuaded to keep them.)

Increasing turnover

There seemed to be a clear role for advertising here. As we have seen, consumers felt that all credit cards were the same and hence they had no particular reason to use one rather than another. We believed advertising had a role to provide reasons why Barclaycard should be used in preference to other cards. If successful this would obviously increase Barclaycard's share of turnover.

Increasing new cardholders

As fees were likely to become a commonplace feature of the credit card market (in fact most major credit cards did go on to charge a fee), it seemed clear that new cardholders would make a more considered choice than had been the case in the past (probably choosing only one credit card). Advertising had a significant role here to persuade potential new cardholders to choose a Barclaycard rather than some other card.

The creative brief

Who are we talking to? What do we want them to do?

  • Existing Barclaycard holders. We want them to keep their Barclaycard and use it more often than other cards.
  • People considering a new credit card. We want them to choose a Barclaycard in preference to other cards.
What do we want to say to them?

Barclaycard is more than just another credit card.

Support

The new Barclaycard benefits (ie purchase protection, etc).

Guidelines

Since 1982 Barclaycard has used Alan Whicker as the perfect vehicle for a strategy focused on Barclaycard's international acceptability. Our new strategy positioning Barclaycard as offering more than other credit cards requires new creative work.

Developing the creative idea

At first, our task seemed fairly straightforward. We simply needed to tell people about the new product to achieve our objectives of defending ownership and growing usage of Barclaycard.

Our initial idea was based on the creative strategy 'all the credit cards you'll ever need', dramatised by a New York ticker tape parade created by people cutting up all their other credit cards. This was accompanied by a voice-over citing the 'new' Barclaycard features. The tone was big, confident and dramatic.

Both Barclaycard and BMP thought this very successfully answered the brief. Our target market however disagreed!

Qualitative pre-testing in June 1990 found significant problems with this approach.

People were reluctant to translate the product claims into a conviction that Barclaycard was better.

This seemed to be because the product features were just verbally listed, not positioned as benefits as a result they tended to get a 'so what?' response.

you'd hardly ever need that
I bet they all do that
but would they really do that when you get down to it?
Source: BMP Research

Interestingly, at this time (October 1990), other research was coming in which supported the pre-testing. It seemed that simply communicating the nature of the different product, was not positioning Barclaycard as a better product.

Awareness of the changes to Barclaycard a result of the letter accompanying the statement sent to cardholders in May, and later direct mail - was high (Table 4).

TABLE 4: PROMPTED AWARENESS OF PRODUCT CHANGES (%)

October 1990
Annual fee93
Purchase protection58
International rescue55
Free Mastercard53
Base: Barclaycard holders
Source: Barclaycard Internal Research

However, there was little evidence that awareness of these changes was improving perceptions of Barclaycard (Table 5).

TABLE 5: BARCLAYCARD IMAGERY

% agreeing with statements
Pre- and post-letter/direct mail
Pre-Post-
April 1990October 1990
Provides special benefits and
services for its cardholders4039
Provides a high quality
customer service4445
Cares for its customers needs4036
Base: Barclaycard holders
Source: Millward Brown

This lack of appreciation of the changes that had been made was reflected in turnover share. The decline in turnover share did not slow down but accelerated (Table 6).

TABLE 6: BARCLAYCARD SHARE OF CREDIT CARD TURNOVER

Pre and post letter/direct mail

PrePost
1989 AverageApril 1990October 1990
*34%34.6%31.6%
* BBA
Source: BMS

Towards a rebrief

It was not as simple as just announcing the new product. People were not perceiving this different product as better.

  1. We needed to talk about benefits not features. To do this we decided to show the features in action. This meant a focus on one at a time.

  2. At the end of the day, the new product features were not going to be needed regularly or by everyone.

    It seemed that turning these extra features into perceptions of a better product demanded adding an emotional dimension to our advertising not trying to rationally persuade cardholders of Barclaycard's superiority.

  3. Therefore it seemed very important that, in contrast to our original idea, the advertising should be positively liked and enjoyed.

A new creative idea

This featured Rowan Atkinson as an opinionated yet bungling secret agent Richard Latham equipped for service with a Barclaycard but continually underestimating how it could help him out of a spot of bother.

This time, pre-testing confirmed that the idea could meet our objectives.

  • The enjoyment and involvement with the idea could sweep aside cynicism, at last allowing people to believe our 'claims' of a superior product.
  • The product features were reassessed as benefits.

This was in marked contrast to the response to the direct mail which, as shown, had successfully communicated the different product, but had clearly achieved little by way of changing perceptions of Barclaycard as being a 'better' card.

Three executions of the new campaign idea were produced in December 1991: 'London' setting up the secret agent theme and restating the international acceptability benefit; 'Moscow' featuring the international rescue service and 'Cairo' featuring purchase protection.

Increasing new cardholders

Barclaycard's share of new credit cardholders rose from 10% pre-advertising to 15% by December 1991. Access share over the same period fell. The ambitious 15% share objective was therefore met and a year ahead of plan.

People getting their first ever credit card are obviously a subset of all new credit cardholders. Barclaycard's share of this 'new generation' of cardholders increased even more significantly after the advertising (Table 7).

TABLE 7: NEW CARDHOLDER SHARE

Pre-advertisingPost-advertising
December 1990December 1991
BarclaycardAccessBarclaycardAccess
Share of new10%23%15%20%
cardholders
Share of first ever15%28%26%28%
cardholders
Source: MAS

Overall, Barclaycard had emerged as turnover brand leader and growing, exceeding all its marketing objectives a year ahead of plan. People were increasingly keeping their Barclaycard rather than other cards, using their Barclaycard rather than other cards, and getting a new Barclaycard rather than other cards.

The effect of this performance on profitability was pronounced. Barclay's CRSDs net loss in 1990 was transformed into a £46 million profit in 1991 (in the midst of one of Britain's longest recessions).

PROVING THE ADVERTISING EFFECT

We have shown how Barclaycard met and exceeded all its marketing objectives.

We believe that advertising played a fundamental part in this. The striking correlations of advertising with marketing success already shown go some way to prove this.

However there are two further ways in which we can prove that advertising was a causal factor in this success:

  • By examining 'soft' data showing how the advertising worked.
  • By eliminating all other variables that might have caused the success.

HOW THE ADVERTISING WORKED

Advertising worked just as we planned.

Advertising awareness

The Latham campaign had considerable impact.

From 13% before the first burst, awareness of Barclaycard advertising reached 34% by the end of 1991.

Access seemed to have adopted a strategy of reducing their investment as a response to the turbulent market conditions (Table 8). As a result, their advertising awareness over the same time period declined from 30% to 20% (source: Millward Brown).

TABLE 8: ADVERTISING TVRs

19901991
Barclaycard7901744
Access1370609
Source: BARB

Importantly, Barclaycard advertising awareness after just one year of the new campaign was nearly three times higher than that of Whicker in its first year (1982), at similar levels of TVRs (Table 9). This demonstrates the success of the new campaign in the difficult task of replacing a memorable advertising property.

TABLE 9: TVRs AND ADVERTISING AWARENESS

ExecutionTVRsPeak and awareness
%
Whicker 1982146912
Latham 1991174434
Source: BARB/Millward Brown

The impact of 'Latham' was such that considerable additional exposure was achieved. (see Appendix).

Product awareness

Awareness of the product changes rose steadily during advertising to exceed the levels achieved by the direct mail communication, (Table 10).

TABLE 10: BARCLAYCARD PRODUCT AWARENESS

% Agreeing with Statement
Pre-1st2nd3rd
AdvertisingBurstBurstBurst
Dec 1990Mar 1991Jul 1991Dec 1991
Provides emergency assistance all over the world37394758
Provides insurance cover on goods bought with the card42576572
Source: Millward Brown
Base: All Barclaycard cardholders

Appeal

The new campaign rapidly established itself as one of the few that really grips the public's attention and affection. Highly appealing advertising was identified as important for success, but is a tough objective in the financial sector.

TABLE 11: ATTITUDES TOWARDS THE CAMPAIGN

LathamAccessAmexWhicker (1989)
%%%%
Enjoy watching it54181525
Like people/characters6391126
It's the sort of advert that67413933
sticks in your mind
Source: Millward Brown 1991

The following quotes are illustrative:

They're just brilliant you always come in and watch them when the wife tells you they're on.
You couldn't get up in the middle of that one.
They're way ahead of the others.
About 1000 times better than the Access ones.
The sort of ad where Barclaycard would gain because people would go to work and talk about it.
Source: Post-testing March 1991

Communication

The advertising communicated our desired product messages extremely well. The Barclaycard feature at the heart of each script came through very strongly (Table 11).

TABLE 11: COMMUNICATION FROM THE CAMPAIGN

'London''Moscow''Cairo'
%%%
World wide acceptance6732
Card replacement-85-
Insurance on goods--72
Others331226
Source: Millward Brown 1991
Base: Adults who had seen the relevant execution

Imagery

Following the advertising breaking, people were beginning to view Barclaycard as not just different, but better (Table 12). This had not occurred after the direct mail.

TABLE 12: BARCLAYCARD IMAGE CHANGES

% Agreeing with statement

Pre-1st2nd3rd
AdvertisingBurstBurstBurst
Dec 1990Mar 1991Jul 1991Dec 1991
Provides special benefits
and services for its card38425966
holders
Provides a high quality
customer service46485051
Cares for its
customers' needs34365052
Source: Millward Brown
Base: All Barclaycard cardholders

In addition, Barclaycard holders were increasingly rating their card more highly than Access holders' view of Access. We had created real 'daylight' between them as planned (Table 13).

TABLE 13: DIFFERENTIATION OF BARCLAYCARD FROM ACCESS

% agreeing with statement
Barclaycard cardholders' view of Barclaycard indexed
against Access holders' view of Access
Pre-AdvertisingPost-Advertising
Dec 1990Dec 1991
Provides special benefits and
services for its cardholders118143
Provides a high quality
customer service112124
Cares for its
customers' needs103152
Source: Millward Brown

Qualitative research confirmed this. We were now successfully positioning Barclaycard as better than its competitors. No longer were people talking about the new product in a disbelieving and cynical way.

It's a better product than Access.
Access is basic compared with Barclaycard.
It's moving away from other cards.
Source: TQC March 1991

The previous tables (Tables 12 and 13) show that Barclaycard holders had an improved perception of the Barclaycard offering and that this improvement in perception was not matched by Access cardholders' view of the Access offering.

Finally, we show that people who had both Access and Barclaycard had increased their preference for Barclaycard following the advertising (Table 14).

TABLE 14: WHICH CREDIT CARD WOULD YOU KEEP IF YOU COULD KEEP ONLY ONE?

Base: Multi credit cardholders (holding at least Barclaycard and Access).

Pre-fee introductionPre-advertisingPost-advertising
Apr 1990Dec 1990Dec 1991
%%%
Barclaycard434059
Access232725
Don't know343316
Source: Millward Brown 1991

After a year of advertising, over twice as many people (who held both a Barclaycard and an Access card) would choose to keep Barclaycard rather than Access if they could only keep one.

People saying 'don't know' were effectively saying that they did not care. We were making people have a preference.

ELIMINATING OTHER VARIABLES

The new product

The single most important variable we need to address is the introduction of the new Barclaycard product.

Although the new features undoubtedly had an effect, it took advertising to gain significant leverage from these product changes.

We have a number of pieces of evidence to support this view

Overall

Firstly, despite the new product being in place, and people knowing about it, Barclaycard's turnover performance between June 1990 and the beginning of advertising in January 1991 showed little movement after the initial fall due to the fee introduction (Table 15).

TABLE 15: BARCLAYCARD SHARE OF TURNOVER 1990 (%)

Pre-feePost direct mail
AprilJuneAugOctDec
34.631.931.531.631.9
Source: BMS

Secondly, simple comparisons of attitudinal data trends with the timing of introducing the new product and the start of advertising, have shown no significant changes in Barclaycard's performance until after advertising (Table 5, 12, 13, 14).

Nothing else changed in January and continued over the course of 1991 to explain why this should be so.

Thirdly, qualitative research had strongly suggested that the product changes alone were unlikely to change peoples' preference for Barclaycard in the way seen. The new features when presented 'raw' were usually met by suspicion and a feeling of 'no big deal'.

Reduction in Barclaycard interest rate

Econometric modelling (see Appendix) has confirmed that relative credit card interest rates bear no relationship to turnover, either for Barclaycard or for the market as a whole. Anyway, any effect would have resulted in share increases between June 1990 and December 1990. This did not occur.

Fee level

Firstly, it could be argued that a higher price in the form of the fee could somehow have raised quality perceptions of Barclaycard. We strongly dismiss this. Extensive research was conducted on the likely response to a fee before its introduction and there was no evidence that a higher price in any way raised perceptions. As we have seen, the reverse was true.

In addition, if this mechanic were to have operated, it would have been evident before the advertising ran. It was not.

Secondly, it could be argued that the fee introductions by five credit cards in 1991, all at £2 to £4 more than Barclaycard (see Appendix) could have accounted for Barclaycard's success in 1991.

However, all evidence suggests that people moving solely on the basis of a fee (the die-hard fee-averse cardholders we have described) will move to non-fee cards, not a card £2 or even £4 cheaper. (This is what happened following the fee introductions before Barclaycard.) By the end of 1991, nine credit cards remained free.

It seems highly unlikely that fee-averse customers would have opted for an £8 Barclaycard in preference to a free card during 1991 if absolute price was the only criterion of choice. If it was not, ie Barclaycard's 'value' over and above price had increased, then we would argue that this itself is an advertising effect. As we have shown, before advertising there is no evidence that Barclaycard was seen as a 'better' product than its competitors.

Other possible influences

There has been no change in any of the following which could theoretically have accounted for Barclaycard's improved performance in 1991 if they had changed.

  • Outlet acceptability of Barclaycard, either absolutely or relatively to competitors.

  • Amount or nature of the direct marketing activity by Barclaycard either to its existing cardholders or to potential new cardholders. (Indeed if anything, recruitment activity by Barclaycard had been cut back, in order to reduce the risk of taking on bad debts, the incidence of which increases in recession.)

  • Profiles, the turnover-based points scheme, running for four years now.

We also know that

  • Barclaycard's response to changes in both mortgage rates and consumer spending is shown by econometric analysis to be almost identical to that of the market, (see appendix for elasticities) and hence could not produce share movements.

  • The decline of Access advertising support in 1991 compared with 1990 did not occur until after October 1991. Not until December 1991, once Barclaycard's pattern of turnover growth was well established, could the year on year decline in Access advertising in 1991 have had any impact on Barclaycard performance, (see Appendix).

So we believe that for the majority of Barclaycard's success in 1991, we have proved that advertising alone was the cause.

Having said that, we have to concede that we cannot prove the role of advertising in one area that is, changing the profile of Barclaycard's cardholders. (We have no competitor data on this and no pre-advertising measure for Barclaycard.)

Many of the non-profitable, fee averse cardholders would have left resulting in a more profitable mix with or without advertising. At best we hoped that advertising could help retain as many profitable cardholders as possible by increasing the 'value' they perceived in Barclaycard.

Although we cannot isolate an advertising effect on the relative proportions of cardholder types, we have, as shown, proved an effect on absolute numbers retained. As such it is difficult to believe that advertising had no effect on helping to retain more profitable cardholders than would otherwise have been achieved.

PAY-BACK ON ADVERTISING

Advertising was planned to generate additional profit in two ways.

Short-term

Firstly, advertising was planned to stimulate increased usage from Barclaycard holders. This increases turnover income (and in addition, interest-based income from the interest paying cardholders).

Econometric analysis has shown that our advertising in 1991 was directly responsible for adding £508 million turnover to Barclaycard, (see Appendix). As a result of the incremental profit this produced, the advertising investment in the short term alone paid for itself more than three times over.

Longer-term

Secondly, advertising was designed to enhance the perceived value of Barclaycard, so that cardholders remain with Barclaycard (particularly important given that new cardholders take some years to become profitable) and new cardholders are attracted. This helps maintain:

  • a large cardholder base needed to deliver economies of scale, most obviously from card processing costs, but there are also other benefits eg negotiations with outlets to accept Barclaycard;
  • resistance to the emergent debit card market;
  • defence against possible future aggressive drives into the UK by foreign credit card issuers.

The longer-term return on this investment is more difficult to directly quantify.

But we can, as an example, calculate that if advertising helped to retain just 2% of Barclaycard's cardholders who otherwise would have been lost, (this estimate is probably an underestimate given the favourable comparisons of Barclaycard losses with other fee charging cards) the profit these would deliver over their 'lifetime' with Barclaycard would more than pay for the 1991 advertising investment.

The real contribution of our advertising to Barclaycard's profitability is therefore significantly higher than 1991 returns alone.

SUMMARY AND CONCLUSIONS

In restrospect, it is easy to underestimate the success of Barclaycard in recovering from the desperate situation they faced back in 1989.

It would have been tempting for them to take the lower risk route adopted by their competitors simply taking more money from their cardholders in the form of a fee, without putting back value into the product through new features and advertising support.

Two principles have been demonstrated which were central to Barclaycard's success.

Firstly, this is not the more typical case study showing how advertising added value to an unchanged product, and hence improved its profitability.

It is difficult to see how Barclaycard could have increased their share of an undifferentiated market with a parity product. A differentiated product was the necessary starting point for growth.

A product with different features was not enough however. Even with knowledge of a different product in place, it took advertising to provide leverage from the changes translating the new product features into valued benefits, in a way which changed attitudes and behaviour and propelled Barclaycard to brand leader.

Secondly, however bad the market conditions, with investment both in the right product and aggressive advertising support, brands can adapt to remain successful and highly profitable. In fact, since these same conditions may encourage competitors to choose to pursue superficially safer strategies by cutting investment, bad times may create opportunities for the brave!

APPENDIX 1: EVALUATION NOTE

The introduction of the optional Barclaycard Mastercard (available only to existing Barclaycard Visa cardholders) as part of the product change in May 1990 is obviously a complication to the evaluation.

To accommodate and remove potential bias from this change we have adopted the following approach to evaluation.

  1. Take-up of the Mastercard option has in fact been relatively modest by the end of 1991 less than 9% of Barclaycard's total turnover was on Mastercard. However, to remove the arguably artificial increase of Barclaycards that the Mastercard introduction has given rise to, all card penetration data (for Barclaycard and competitors) is presented in terms of cardholders not cards. This removes the effect of the fact that an (albeit) small proportion of Barclaycard holders have two sorts of Barclaycard.

  2. We have evidence, both qualitative and quantitative, that Barclaycard holders who have opted for Barclaycard Mastercard in addition to Barclaycard Visa are using the extra card simply (as intended) as a means of re-organising their spending on Barclaycard across two cards instead of one.

Because of this, and to simplify the evaluation, Barclaycard turnover is analysed as total brand turnover, not split by the two variants. This seems appropriate anyway given that Barclaycard Mastercard holders are acquiring the extra card because of and through the Barclaycard brand, not the network organisation. (Most have no idea what Mastercard is or means.)

The introduction of Barclaycard Mastercard cannot have been a factor responsible for the changes in Barclaycard performance in 1991 for the following reasons:

  1. Barclaycard Mastercard was available from May 1990. Nothing has happened during 1991 in terms of recruitment or usage directed activity which could have changed Mastercard behaviour in a way which would have turned round the performance of the Barclaycard brand as we have seen.

  2. Econometric modelling which was undertaken just on Barclaycard Visa turnover has indicated an advertising effect on Barclaycard Visa turnover which, as shown, alone (ie without Mastercard turnover) paid for advertising more than three times over.

APPENDIX 2

Credit card fee introductions

£
Save & Prosper8March 1989
Lloyds Access12December 1989
Barclaycard8May 1990
(Visa/Mastercard)
Midland10February 1991
(Access/Visa)
Firstdirect10May 1991
Bank of Scotland10July 1991
(Visa/Mastercard)
Halifax10July 1991
NatWest12September 1991
(Access/Visa)

Non fee charging cards at December 1991

TSB Trustcard
Royal Bank of Scotland
Clydesdale Bank
Co-operative Bank
Bank of Scotland
National & Provincial Building Society
Leeds Building Society
Yorkshire Bank
Town & Country Building Society

APPENDIX 3

The Richard Latham campaign was featured in seven national newspaper articles and one double page article in the TV Times during 1991.

We estimate that these have contributed in the region of £50,000 of additional, free media space.

The campaign has also won the following awards:

D & AD 1991CategoryAward
Moscow, CairoTelevisionNOMINATION
MoscowTV Commercial up to 60'NOMINATION
Cannes Film Festival 1991
MoscowFinancial CategoryGOLD
CairoFinancial CategorySILVER
EPICA 1991
MoscowFinancial ServicesFILM WINNER

APPENDIX 4: RESULTS OF THE ECONOMETRICS

Explanation of how economic factors and advertising impact on Barclaycard

This section explains in more detail how turnover is affected by credit card interest rates, mortgage rates, consumer spending and advertising. It also explains how the overall increase in Barclaycard turnover attributed to advertising was calculated.

The effect of credit card interest rates

As part of the econometric analysis the relationship between turnover and a range of credit card interest rates was examined. These included Barclaycard, Lloyds Access, NatWest Access, Midland Access and a number of smaller issuers.

The absolute level of credit card interest rates had no statistically significant effect on turnover at total market level or on Barclaycard. Neither was Barclaycard affected by the relative level of its interest rate versus competitors.

This was in fact consistent with theory and prior expectations. Qualitative research shows little consumer knowledge of interest rates. Barclaycard's decision to lower interest rates had anyway been based on a fairer pricing policy (full payers v extended credit users) and not on the expectation of increasing turnover.

The effect of mortgage rates

Mortgage rates do directly affect consumers' willingness to use a credit card as a means of payment when shopping. (This is an effect over and above any effect of interest rates on the level of consumer spending as a whole; that is picked up by the model by retail sales trends.)

A decrease in the mortgage rate will, all else remaining equal, cause an increase in credit card turnover. Any effect of mortgage rate changes is however short-lived; a decrease will only boost turnover in the month when interest rates change after which turnover returns to normal.

The effects of mortgage rate changes are almost identical for Barclaycard and the market and thus a mortgage rate change has no implications for Barclaycard share.

As we do not wish to divulge the exact mortgage rate elasticities, the elasticities are presented in relative terms, not absolute numbers (Barclaycard mortgage rate elasticity = 1.00, total market = 1.05).

The effect of consumer spending

Interpolation of the CSO quarterly consumer spending series to make it suitable for a monthly analysis proved unsatisfactory and so retail sales volumes have been used as a proxy for total consumer spending.

Barclaycard's response to the recession and the slowdown in consumer spending is almost identical to that of the market as a whole, so again there are no implications for Barclaycard's share (relative elasticities, Barclaycard = 1.0, total market = 0.99).

Calculating the contribution of advertising

This was done by model simulation. Although the simulation is based on data up to August 1991 and thus only covers the first two bursts of advertising, the Barclaycard model continues to track the data well between September and December and thus it is legitimate to extrapolate the effect measured in the first and second bursts of advertising, to the third. The results of the simulation estimated the overall effect of the first two bursts (a total of 1334 TVRs) to be £297 million ie £22.26 million/100 TVRs. Thus the total incremental turnover generated in 1991 by 1744 TVRs = £388.28 million (at 1987 prices).

Translated to current prices this is an increase of £508 million.

The effect of competitive TVRs

The overall weight of Access advertising up to June 1991 at 609 TVRs was slightly above that in 1990 (502 TVRs) and thus if anything would have been a depressing factor on Barclaycard share. The only month where the differing pattern of Access advertising could have benefited Barclaycard was December (the effect of Access advertising is seen one month after airtime), once Barclaycard's pattern of turnover growth was well-established.

BARCLAYCARD AND ACCESS TVRs IN 1990 and 1991 (UNWEIGHTED)

BarclaycardAccess
1990199119901991
January4780210
February446000
March026135072
April3080131333
May000196
June045208
July016100
August00150
September0040
October0000
November03594690
December0513800
79017441370609

APPENDIX 5: TECHNICAL APPENDIX COVERING THE ECONOMETRIC ANALYSIS

The econometric results cited in this paper have come from a consultancy project carried out for Barclaycard in 1991. The results are confidential to Barclaycard and it would be inappropriate to produce the full results or the full technical specifications of the models here.

Included below however is a detailed account of how the models were produced, indicating the variables which were investigated during the course of the analysis and showing the technical validations for the models.

Models were produced for total credit card turnover and for Barclaycard Visa turnover. (At the time the project was started, the data on Barclaycard Mastercard was considered too limited to be modelled specifically.) Having a model of the market and of Barclaycard Visa turnover allows implications for Barclaycard Visa market share to be derived.

The analysis investigated the impact of:

  1. Economic factors such as consumer spending, retail sales, interest rates, mortgage rates and consumer incomes

  2. Market specific factors such as numbers of cardholders, credit card interest rates, fee introductions and advertising both by the main players and by a number of smaller advertisers. Although not strictly part of the market, we did investigate the impact of American Express activity too.

  3. Barclaycard marketing initiatives such as profiles and numbers of mailings. The turnover data used was the monthly data produced by the British Bankers Association (BBA) which is considered to be the most reliable long-term series on turnover. The data used covered the period from April 1986 when the BBA data first becomes available to August 1991. The models were thus based on 65 data points. The differences in numbers of data points reflected in the test statistics for the separate models relate to both the slightly differing lag structures of each model and the degrees of freedom.

The BBA was also the source of the data relating to credit card interest rates and other market specific factors.

Other data sources were:

  1. The CSO for retail sales, consumers' expenditure, mortgage rates, base rates
  2. BARB and Media Register for data on advertising activity.

The market is seasonal with low points in February and March and a seasonal peak in December. This seasonality tends to be regular from year to year and does not depend on the behaviour of a secondary variable eg temperature. It was therefore captured using seasonal dummies.

The data indicated that log specifications were appropriate for both models. Both were rigorously tested and the results of the tests are produced below. All the coefficients had signs and values consistent with theory and there was of course the opportunity to cross-check the results of both models with each other as a further validation.

All variables included in the market model had t-statistics in excess of 2.5 with the exception of one of the lags on Access advertising (t-stat = 1.67). In the Barclaycard Visa turnover model, all but two variables had t-statistics in excess of 2.1. The remaining two had t-stats of 1.97 and 1.56 respectively. Statistical testing indicated however that the models would be mis-specified if these variables were dropped.

Neither real credit card nor real Barclaycard Visa data consistently under- or over-predict turnover when actual and fitted data are compared.

RESULTS OF STATISTICAL TESTS

Total credit card turnover model

The following tests for fit and specification were carried out and the model of total credit card turnover performed very satisfactorily against them all.

0.990
_
0.987
Durbin-Watson2.12
Durbin's h testnot applicable
Standard Error of Regression (log model)0.0232
Variance Ratio Test1.1085
F(20,22) = 2.07
CHOW Parameter Stability0.3789
F(13,42) = 1.96
CHOW Post Sample Predictive Failure1.0111
F(5,47) = 2.40
LM 1 (Critical value = 3.89)1.6738
LM 12 (Critical value = 21.0)14.0695
ARCH 1 (Critical value = 3.89)0.2612
Skewness and Kurtosis0.1700
(Critical value = 5.99)

Barclaycard Visa turnover

The test statistics for the Barclaycard Visa turnover model were as follows:

0.957
_
0.942
Durbin-Watson2.044
Durbin's h test0.2028
Standard Error of Regression (log model)0.030
Variance Ratio Test1.278
F(19,15) = 2.34
CHOW Parameter Stability1.2512
F(17,34) = 2.08
CHOW Post Sample Predictive Failure0.3996
F(5,41) = 2.33
LM 1 (Critical value = 3.89)0.3280
LM 12 (Critical value = 21.0)16.81
ARCH 1 (Critical value = 3.89)0.5020
Skewness and Kurtosis0.4072
(Critical value = 5.99)

NOTES & EXHIBITS


Barclaycard: How a Bungling Secret Agent did more than you'd credit

FIGURE 1: CREDIT CARD MARKET GROWTH (INDEXED ON 1974 = 100)

Source: BBA