Research: Words' worth

If word of mouth is anything to go by, Ocado appears the obvious
choice.

Tales of the quality of its goods and services have spread from one
satisfied customer to another. Hardly surprising, really, given that
this formed an integral part of its strategy.

'Its whole ethos was quality over coverage,' says Giles Hedges, joint
head of planning at Miles Calcraft Briginshaw Duffy. 'Quality of
customer experience first, coverage second. It expanded from postcode to
postcode, gaining consumers in smaller quantities than its rivals, but
in concentrated areas.'

Clearly, putting word of mouth at the front of a strategy to communicate
with customers can be a valuable exercise. Yet for marketers, building a
word-of-mouth campaign is in many ways the easy part; measuring its
effectiveness is a different matter entirely. There are similarities to
the early days of sponsorship, when brand owners basked in the reflected
glow from association with a winning team, but were often unable to
isolate the effectiveness of their spend in this area compared with
other elements of the marketing mix.

There are specialist research companies - as well as relevant software -
to examine the most important metrics, such as spread, analyse the tone
of resulting consumer conversations, identify the key connectors and
work out how a campaign is affecting a company's image. There is even a
Word of Mouth Marketing Association in the US that aims to create
accountability in this area.

Yet still there are no internationally recognised standards.

However, there is now movement toward achieving this goal. London School
of Economics' Dr Paul Marsden will reveal new evidence at the Market
Research Society conference in March that shows how research outreach
programmes can be used to harness word-of-mouth connections.

Power of opinion

Marsden, whose background is in pharmaceuticals marketing, points out
that one of the first research studies on word of mouth took place in
this sector. 'Doctors are overloaded with ads and information,' he
says.

'They don't take notice of interactive marketing, so when you're
launching a drug that cost a quarter of a billion dollars to develop
it's vital that you get customer advocacy, because that drives
sales.'

His attention was drawn to parallels between pharmaceuticals and FMCG
marketing, in which information overload and fragmented channels result
in consumers taking insufficient notice of advertising. When a study on
the topic appeared last year in the Harvard Business Review, he decided
to replicate it to see whether its findings were valid in the UK and
whether a monetary value could be put on that word-of-mouth metric.

'We use a Net Promoter Score (NPS) system used by Bain Consulting among
others,' he says. 'It is correlated to business growth. You ask one
simple question - how likely is that you would recommend whatever to
whomever - on a scale of 0 to 10. You take all the people who say nine
and 10, subtract all those who answer 0 to six, and you get the net
promoter score.

That score, based on that one question, is predictive of sales growth.'
Customer satisfaction, he says, has no correlation to future sales and
is, therefore, not particularly good as a key performance indicator, in
contrast to word of mouth.

Industry discussion on the topic has to date been divided, as deciding
on a consensus of how word of mouth works has proved difficult. This was
acknowledged in a report on the topic by Mediaedge:cia's think-tank,
MediaLab. Called 'Where's Debbie', its introduction explained that word
of mouth 'remains an enigma for marketers because intuitively accessible
theories on how it works do not connect with research methodologies that
allow for accountable planning, implementation and evaluation of
word-of-mouth strategies'.

This is echoed by Martin Oxley, managing director of interactive at TNS,
who says that 'as a concept, it is very powerful, but as something
measurable, word of mouth has always struggled. How do we get it on a
measurable footing?'

Gauging impact

It is a problem Oxley and his colleagues are determined to solve,
particularly in light of the shift toward non-traditional advertising.
'We are finding that when launching products many companies are not
necessarily advertising.

Heinz, for example, has more or less stopped it and said it will go
below-the-line.

So we are trying to show that there is this other aspect companies need
to measure: the extent to which the diffusion of innovation can be
influenced by the types of people buying your product, the extent to
which they will buy the ideas and their connectedness.'

It is small wonder that companies are shifting budget when studies such
as one conducted last year by GfK NOP indicate that 92% of US consumers
now view word of mouth as the best source of ideas and information about
new products, compared with only 67% in 1977.

The difficulty lies in predicting precisely which consumer types are the
ones to watch. There are a number of theories concerning the profile,
distribution and value of influencers. Popular wisdom has it that one in
10 people are influencers and they tell the other nine what to do. The
reality, as David Evans, research director at Continental Research,
notes, is more complex. 'A number of studies that we have carried out
have shown that certain people are more important than others with
regards to word of mouth,' he says. 'This, however, varies depending on
the product sector analysed. For example, someone who is happy to
communicate the benefits of an iPod to others may be less likely to do
the same regarding a Dyson product and vice versa. Segmentation,
therefore, is vital to draw out the key segments within the sector which
are most likely to pass on information.'

Generating influence

It can also prove a challenge for companies to make sense of the stream
of information resulting from a word-of-mouth campaign. The sheer scale
of such operations is evident at Procter & Gamble, whose Tremor division
recruited 280,000 teen 'influencers' in the US who now spread the word
for P&G and third-party clients on anything from shampoo to movies.

'The wealth of consumer-generated qualitative media that is often
created means that quantification is necessary to enable accurate
analysis and insightful interpretation,' says Amanda Scott, a research
manager at FreshMinds.

'Word-of-mouth researchers should mix quantitative digital activity data
and mining of digital archives with qualitative data from consumer
surveys and testimonials to give a rounded picture of the impact of a
word-of-mouth campaign.'

However, not all researchers are happy with an overwhelming focus on
quantitative data in the discipline. 'If all you can produce is
quantitative statistics about word of mouth, all you are doing is saying
this is very important without being able to do anything about it,' says
John Griffiths, who runs the Planning Above and Beyond website for
account planners. 'The question should be: are some people better at
word of mouth, and if so, why? Is it because their whole status and
social position is predicated by how effective they are at always
telling people things and knowing the right people?'

Clients will no doubt discover much from Marsden's session. After all,
which marketer would not want to know about a technique that could
predict sales growth? Yet there is a danger that such a focus ignores
the real intricacies of human behaviour: the whys and wherefores. Surely
no marketer could resist a technique that could identify what makes a
word-of-mouth story more interesting, and the person who tells it more
persuasive than the rest?

CASE STUDY - DISCOVERY CHANNEL

Discovery Channel commissioned a study into the influences on people's
choice of holiday destination in three European markets. The aim was to
help communicate to tourism authorities the benefits of using its
channel to reach their target audiences.

A key aspect of the holiday decision-making process is word of mouth,
yet conventional tracking often deems such recommendations a 'dead'
figure, preferring more concrete sources of information.

Continental Research used a statistical technique, the Markov process,
within a conventional segmentation study. This traced back influences on
choice of destination to find the ultimate source of advice for each
respondent. The Markov process worked out how people who passed on the
information were influenced down the chain, using statistical assumption
to provide a total impact for each medium, from brochures and weekend
supplements to TV shows and personal recommendation.

Discovery demonstrated that its viewers were more likely to be
influenced by programming and communication about holiday destinations.
Forty per cent of viewers were influenced directly by programmes and 27%
indirectly (such as when someone who has been influenced by a TV
programme influenced someone else by word of mouth). They are also more
likely to pass this information on to others than non-viewers.

The findings were presented at last November's World Travel Market, and
helped to secure new advertising budgets.

Research: 2006 Word of mouth
Date: Thursday 23 March
Venue: Barbican, London
Contact: www.mrs.org.uk

Details: Chaired by Mark Earls, planning director at Ogilvy, the session
will be in three parts.

1Simon Chadwick, partner at Cambiar, and Ed Keller, co-author of The
Influentials, will report on experimental research designed to
investigate whether all individuals are equal in their peer
influence.

2Dr Paul Marsden of London School of Economics will present new evidence
from a major FMCG client that quantifies the effect of word of mouth on
business performance.

3Graham Trayner, a director of Opinion Leader Research, will reveal how
research can turn passive consumers into active purchasers.