Why does everyone hate Millward Brown? - Pre-testing of ads has become the most heated area of debate in the business. Damian Lanigan asks if the prime suspect, Millward Brown, has a case to answer

A quantitative research company based in a business park near Leamington Spa. Could there be anything more tedious? Yet the very mention of Millward Brown is enough to send some advertising people into a slavering rage.

A quantitative research company based in a business park near

Leamington Spa. Could there be anything more tedious? Yet the very

mention of Millward Brown is enough to send some advertising people into

a slavering rage.

What does it do that gets people so worked up? Burn babies or


Well, in the eyes of many in ad agencies, particularly creatives, this

is exactly what it does. It takes a new-born, fragile advertising idea

into a room, subjects it to torture with long questionnaires and a

joystick and then returns the mangled remains to the agency,

recommending that the sorry item is put out of its misery.

Millward Brown rejects this caricature. It has supporters in the client

community, and some in agencies, but the image of ’ad killers’ persists.

This is despite the fact that the Link Test - Millward Brown’s

pre-testing technique - accounts for a relatively small part of its

revenue. So is it possible to start a debate from a more constructive


As one of WPP’s six market research subsidiaries, Millward Brown is

undoubtedly a successful organisation whose ideologies have retained

great power within UK marketing practice. Martin Sorrell, WPP group

chief executive, once described Millward Brown as ’the jewel in WPP’s

crown’ and most agencies and clients have worked with it through

tracking studies or pre-testing techniques. In fact, it is the very

strength of Millward Brown’s market position that makes it the target

for criticism of all quantitative advertising research. But there is

praise as well as vilification.

The work of Gordon Brown, the retired co-founder, on advertising

effectiveness was instrumental in forming notions of how advertising

works. In the words of MT Rainey, planning partner of Rainey Kelly

Campbell Roalfe: ’Millward Brown’s work in aiding our understanding of

advertising through tracking has been a great asset.

It played a key role in creating the research culture in this country

that has always supported the creative imperative.’

In particular, she values Millward Brown’s historical focus on the

importance of cut-through: ’The identification of creativity as the most

important variable in generating awareness certainly contributed to

advertising practice in this country.’ However, the Awareness Index,

which is Millward Brown’s means of recording the efficiency of a given

ad at generating awareness, does not always provoke such positive


Dan O’Donoghue, chief executive of Publicis, comments: ’The Awareness

Index lulls people into a false sense of security - ’it’s done really

well on Millward Brown, it must be all right’. In general I don’t like

Millward Brown’s approach because it seems to be rooted in old-fashioned

weight theory.’

He adds: ’There are new uses of media that are elusive to being

tested ... we’ve used a media rotation strategy on the Renault Clio, but

you just can’t predict the effects of this using Millward Brown.’

But any criticisms levelled at the Awareness Index and tracking in

general are insignificant compared with those reserved for Millward

Brown’s Link Test, which costs upwards of pounds 12,000 for one ad to be

pre-tested by an easy-to-reach sample of, say, 150 housewives.

MT Rainey quotes Alan Hedges’ 1974 seminal book on pre-testing, Testing

to Destruction: ’ ’It is not possible to make a realistic test of the

effectiveness of a commercial in a laboratory situation in advance of

real-life exposure. Until this simple but uncomfortable truth is grasped

much advertising research will go on being sterile and unproductive.’ It

is axiomatic that you can’t quantitatively pre-test ads adequately.’

O’Donoghue is more blunt: ’I don’t believe in quantitative testing.

There is no satisfactory methodology. You just can’t do it.’

The reasons cited for lack of faith in pre-testing are many. First,

there is the belief that it is impossible to replicate the real world in

a laboratory.

Beth Barry, the planning director of Ogilvy and Mather, says: ’What

Millward Brown (and everyone else in pre-testing) has yet to get over is

the old problem of the ’one or two exposures to an ad’ being

representative of a whole campaign. Similarly, there is the issue of

animatics and how well they can project the effects of a finished


O’Donoghue adds: ’Pre-testing is predicated on the total denial of

creative craft skills and the difference they can make to an ad.’

Rosi Ware is Millward Brown’s managing director - since Brown’s

retirement in 1992, the public face of the organisation. How does she

respond to such fundamental criticisms?

’We agree that it is impossible to replicate real life, and we’ve never

claimed that Link Test does that. It is based on the premise that people

are selective about what they retain. Link Test identifies what people

respond to and remember and gives an indication of whether it (the ad)

is going to grab people’s attention. This is why we ask for a ’stream of

consciousness’ response in the test. We ask people to respond as if they

are talking about the ad down the pub.’

And what about the contention that the animatic Link Test cannot account

for the contribution that the director, music track and casting make to

a finished film?

’Again, I agree that animatics can’t replicate the effects of a finished

film. But if the stimulus material is good then the diagnostic measures

within Link Test can help develop a style, and give steers on casting

and music. The key is the quality of the stimulus material.’

According to Rainey, the increasing tendency for advertising to play a

bit part in an integrated marketing mix also undermines the usefulness

of pre-tests: ’There is a move to high-concept marketing ideas rather

than just ads. A campaign such as Miller Time cannot be tested using

conventional pre-testing methods.’

Ware doesn’t disagree: ’Link Test can’t help to evaluate the total

campaign mix. A good example is the Mondeo launch which made the News at

Ten. Link can’t account for that type of PR. However, I would encourage

clients to Link Test all executions in a new campaign, rather than just

one or two. We acknowledge that a campaign can add up to more than the

sum of its parts.’

Even when confined to the conventional remit of ads, many believe that

Millward Brown’s techniques are inflexible. Barry says: ’There are

questions attached to the perceived rigidity of some of Millward Brown’s

branded products which need more sensitivity on both sides. For

instance, the notion that advertising always works in a linear, logical

fashion, which isn’t true.’

Millward Brown’s bid to project tracking performance from Link Test is

also a source of concern. O’Donoghue says: ’Millward Brown has shot

itself in the foot. The attempt to connect Link Test to the tracker is

to create an increasing circle of irrelevance to the real world.’

Ware does disagree on this point: ’Link Test was based on 15 years of

tracking studies, and their observations on which ads tended to perform

better in the marketplace. Also, Link Test gets better all the time. I

think strong negative views are from people who don’t have recent

first-hand experience. The bad experiences people have had tend to


Crucially, Barry sees an issue that is as much the fault of agencies and

clients as Millward Brown: ’There is a problem concerning how we all

interpret the data we are given. Perception of Millward Brown’s branded

products overwhelms understanding of their usefulness.

Like all research, what Millward Brown offers are not finite products -

you get out what you put in.’

Rainey agrees that Millward Brown’s results are used badly: ’They tend

to be used as a binary decision-making tool. Clients take them too

literally.’ However, she attributes some blame to the way in which

Millward Brown presents its findings: ’Too often the results are

presented by people who do not understand the implications of what

they’re saying. These results can make or break agency/client


On the calibre of people, Ware contends that Millward Brown is making

progress: ’We’re a people business and our people have different

qualities and abilities. We try hard to match client and agency teams,

and now make our trainees spend time in agencies so that they understand

how awful it can be to have your idea researched.’

Ware’s advice to agency people who have a Link Test approaching is

clear: ’Talk to us and make sure we’re part of the process. I try to

ensure that everybody involved on a project here looks at the creative

brief and understands what the advertising is trying to do.’

One area where there is complete agreement is that Link Test is a vast

improvement on US-style ’persuasion shift’ testing. These tests are

predicated on one-number results - ’if it’s a seven, run it; if it’s a

six, kill it’, and they are gaining ground in the UK. Compared with such

brutal methods, Link Test begins to look like something of a sweetie. As

Barry says: ’Believe it or not, after any time spent in the US you want

to rush home and give Rosi a cuddle.’ Steady on.