Direct marketing: Keep it clean

antagonisation of existing customers.

With a multitude of data being generated from myriad sources, keeping on
top of it all is demanding. It is estimated that 10% to 20% of mailing
lists are out of date. And if lists are not regularly cleaned and
updated, effectiveness plummets.

'Some lists will have a higher percentage of old data than others,' says
Scott Logie, managing director of data strategy firm Occam, and deputy
chairman of the DMA Data Council. 'It depends on where and when the list
is sourced, how often it is cleaned, and if it is regularly suppressed
and updated.' It is vital not to assume a list is perfect. 'People move
house every day; no list will ever be spot-on,' Logie adds. 'And people
change their habits, opinions and purchasing decisions regularly.'

Occam encounters many lists that are assumed to be clean but which
actually contain poor-quality data. However, he claims it is rare that
such lists cannot be cleaned in order for them to be used without fear
of inflicting brand damage.

Data collected via the internet has many advantages. It allows clients
to take control of their own databases, helping to reduce their reliance
on lists. If mined more deeply, it offers rich data to marketers.
'People have changed their buying behaviour. More and more people are
going to the internet to compare costs, sizes and styles,' says John
Wallinger, database director at Craik Jones Watson Mitchell Voelkel.
'Using the web as a hand-raising tool is important. We use lists, but
less so. Budgets are moving online, and the audience is warmer
there.'

Appealing to this 'warmer' audience is the approach preferred by Honda
in the UK. In the past, it has used cold lists based on surveys, but it
now prefers not to use lists for cold prospecting wherever possible.
Even with surveys that ask 'When do you intend to change your car?',
Honda has found responses at odds with reality.

Neil Fretwell, data director at relationship marketing agency Hicklin
Slade & Partners, manages Honda's prospects and customer database, which
includes everyone who has made contact with the company via a list or an
enquiry. '(We write to prospects) regularly to check and recheck that
the information we hold is accurate,' Fretwell explains. 'We will know
if they are driving the same car as six months ago, and whether they
still intend to change their car in the next six months.'

Like many brands, Honda uses digital channels to update its information,
finding it cost-effective and efficient, with email the preferred
contact method. Mick Doyle, Honda's manager of customer understanding,
says there is a focus on obtaining as many email addresses as possible
through its dealership network. Information is kept on a central
database and all communication is coded.

The car maker prioritises two questions: when are you going to change
your car, and what model are you currently driving? Fretwell says this
simple approach is useful in a data-heavy world where 'far too many
brands waste time chasing the more irrelevant bits'.

Yet for all the advantages of online data collection, there is also a
downside. Logie argues that consumers' use of 'spoof names' is on the
rise, in part a backlash against excessive online form-filling. This
should be taken into account when formulating data-capture plans. Logie
adds that many in the data industry need to be 'more intelligent' at
formulating means of identifying web users. For example, the capture of
date-of-birth information could be prioritised via a pop-up window.
However, he warns that it is important not to overburden consumers with
too many questions, which could prompt them to decline forms or falsify
their entries.

Failure to get consumers' details correct can do more damage than just
annoying the erroneous recipient. In the era of the proactive customer,
with rapid feedback via blogs and product-review sites, brand perception
is being driven more than ever by word-of mouth advocacy. 'If data is
not comprehensively updated and cleaned, then the negative impact on a
company's brand - through contacting the wrong consumers - can erase the
financial effect of an otherwise high-performing campaign,' says Mark
Challinor, head of data planning at Archibald Ingall Stretton.

Pooled data - the buying-in of a number of lists to bolster the volume
of information used - has proved popular in recent years, but some
industry watchers are concerned that there is no obligation to disclose
the original source of the data. As a result, companies may face the
risk of buying something they already get from another supplier.

'Pools require greater transparency, and it is something I have been
urging the DMA Data Council to act upon,' says Lorna Keane, list broking
account director at Prospect Swetenhams. 'The industry's reputation has
been hit hard recently, and it's far too risky to be sending out
mailings or emails using data that is potentially of poor quality, and
which may not even have been (measured) against industry standards.'

Although list brokers and data owners have to comply with legislation
such as the Data Protection Act, and the suppression of the Mailing
Preference Service, they are effectively self-regulated when it comes to
data performance, quality and accuracy. Challinor is among those who
feel this system has worked relatively well for the industry, as those
companies that invest in database maintenance, updating and quality
improvement are the ones that perform well for their clients, meaning
they are retained as suppliers. Those companies that fail to invest in
such measures will, he argues, invariably be dropped.

Nevertheless, marketers do not want to learn the hard way that their
list is flawed, and there are calls from within the industry for an
audited quality standard for consumer lists. This has already happened
in the B2B sector, where after a tentative start the DMA Business List
Audit scheme is gaining an aura of authority. The audit comprises an
online accreditation process, plus a site visit from a member of the DMA
compliance team and random data checks. Annual renewal involves a fresh
audit, which includes a customer-satisfaction survey.

'These are big steps forward to audit data properly so that it does what
it says on the tin in relation to timeliness and accuracy,' says Michael
Howe, head of 100 Percent Direct Marketing. 'The audit is a level of
baseline quality that can be trusted. There are lists out there that are
particularly bad and will never get a certificate.'

The DMA Data Council has discussed the introduction of a similar system
for the consumer sector, but no concrete proposals have yet been made,
perhaps because the scale of the task is daunting.

Yet for brands themselves, reaching out to potential and existing
customers through data capture can also be intimidating - the amount of
information being created is vast, and the methods of capture various.
So how can brands ensure, for instance, that they have a holistic view
of their consumers?

It is a challenge that Matthew Button, Lexus GB's CRM and database
marketing manager, is all too aware of. Using its own 'segmentation and
propensity' model, while working with agency Partners Andrew Aldridge,
Lexus assesses the likelihood that a potential customer will buy one of
its cars, and applies an 'emotional index'. Combining its own data with
the geodemographic methods employed by Mosaic and TGI, Lexus segments
customers into those likely to be swayed by emotional brand factors such
as value and aspiration, and those more motivated by rational
considerations such as finance deals and running costs. The company's
campaigns are then tailored to these groups accordingly.

'The way we cut up the data varies from campaign to campaign,' explains
Button. 'The job is about continually refining whatever insight you
might have. If you have some kind of theory about your customers'
behaviour, it's your duty to validate it.'

In March, data heavyweight Experian announced it is to merge its
ClarityBlue, Prospect Targeting, Bureau and Intact businesses to create
a single entity known as Experian Integrated Marketing. Providing
technology, data, consultation and analytics, the new brand highlights
how demand for data insight and services is growing and evolving. The
move is also an illustration of how tired old lists and superficial
segmentation will no longer do.

CASE STUDY - NECTAR

Loyalty Management UK's (LMUK) Nectar programme, launched in 2002,
generates a mountain of data. Roughly half of all UK households have the
card, which can be used in about 6000 retail outlets.

In March, Sainsbury's announced that it had renewed its contract with
the loyalty group and will benefit from LMUK's new data analytics
business, established to provide greater insight into customer shopping
trends, gained from analysis of till and Nectar information.

LMUK segments its active customer base into seven levels of commitment,
ranging from 'engaged collectors' to those who are 'passively swiping'.
Personal information supplied by members when registering for the scheme
is regularly overlaid with Experian Mosaic data to provide insight into
life stages and help highlight targeting opportunities. 'It's relatively
easy for organisations to collect a lot of data,' says Nectar marketing
director John Sheekey. 'It can be hard to extract meaningful information
out of that data, although it becomes easier as you get to know your
database better.'

Sheekey adds that Nectar is moving toward more 'trigger-based'
communications, with consumer behaviour identified from data changes
sparking action. An example is the 'home mover' pack sent out to Nectar
cardholders when they change address. This includes a map with the
cardholder's new home featured, indicating where Nectar points can be
collected locally. Coupons are included to help track impact.

Data is also scrutinised to identify members with an interest in green
or sustainable products. Such products are targeted at segments likely
to be receptive using tailored campaigns.