Special Report: Email - it's about quality

Fast forward to the present: an estimated 800 million emails were sent
in the first three quarters of last year, covering a variety of
activities, according to the Direct Marketing Association's Email
Marketing Council's National Email Benchmarking Survey (see chart,
below). Mailing volumes also increased by 79 per cent in the period
between October 2004 and September 2005.

Combine these facts with a decline in bounce rates and spam complaints
as well as an increase in clickthroughs and it all adds up to one thing:
evidence that email marketing is on the up and proving its worth as a
direct marketing channel. The speed, low cost and flexibility of email
have made it an attractive medium and companies from all sectors are now
using it to communicate with their customers and prospects.

But while emailing volumes are on the increase, there is concern in some
quarters that email data quality is being overlooked. So are these
concerns justified?

There are two main areas of contention. The first concerns the richness
of the email data available.

"More than eight years after the introduction of email data, too few
list owners segment their offering with demographic or behavioural
selections.

The result is an inability to target effectively across most files,"
says Claire Hart, a director of direct marketing consultancy Caspian
Partnership.

"Targeting is not a top priority for clients, simply because the cost of
email data and the associated campaign overheads of getting it out of
the door are so low. If email marketing carried the same costs as direct
mail, I believe the market would have demanded more in this
respect."

As Hart points out, there is some rich email data out there. IPT, for
instance, has created high-volume, data-rich email databases from
information gathered by websites that reward consumers in different ways
for opting into third-party email communications.

"But," says Hart, "promotionally led data collection methods can draw
audiences that are more interested in giveaways than making active
purchases."

Rich, highly profiled email prospect lists may be in short supply but
there is no reason why companies should not have this type of data
available for their customers. However, too many companies do not use
the customer data they hold to segment their email communications.

Poor targeting is the most common mistake. For brands, it can be all too
easy to send the email to everyone on the list, creating lazy
practices.

Then, when brands do have something relevant to a particular consumer,
they have either unsubscribed or they hit the delete button because they
have been over-marketed to in too generic a way.

Nick Fraser, director of business-to-business (B2B) marketing at
Experian, shares this view. "In some ways, email marketing is too easy
for its own good," he says. "Really, it's no different to direct mail.
You have to segment the database and use specific messaging that is
relevant to each segment."

Data accuracy

The other big issue in email data quality is the accuracy of the
information.

Compared with postal data, email information is highly volatile, with
consumers changing service providers, setting up new web-based accounts,
running multiple accounts simultaneously and letting other ones lapse,
all without a thought for the direct marketers who are trying to target
them.

"There is no equivalent of the National Change of Address file for
email," says Chris Duncan, managing director of data management business
Alchemetrics.

"You can sometimes bridge it via mobile numbers or names and addresses,
but they are just commodities. It is a bit hit-and-miss compared with
postal data."

When email service providers receive an undeliverable email, all they
can do is establish why it bounced back so they can delete it from the
database if it is a dead address, or amend the address if it has
changed.

Denise Cox, newsletter specialist at email newsletter producer
Newsweaver, accepts that email data decay is inevitable. But, she says,
it is not impossible to keep information refreshed in the B2B world.

"If it is manageable and the value of the addresses merits it, consider
contacting hard bounces by phone or post to get a new address," she
says.

If you are buying in B2B data, this probably already happens as a matter
of course. Experian's email data, for example, is bought in from Thomson
Directories, which updates the data annually.

In B2B, however, things are more complicated. It is a question of
looking at the emails that are not delivered each time a campaign is run
and assessing the cause of the problem.

Neil Anderson, managing director of campaign software provider Neolane,
says the delivery rate is typically quite low when new clients start
sending emails to their databases because a large proportion of the
addresses have not been mailed on a regular basis and are likely to be
out of date.

"But after a few campaigns, when the bouncebacks are fed into the
software and quarantined each time, the data is gradually cleansed and
the chance of delivery reaches as high as 98 or 99 per cent," Anderson
says.

Calls for a kitemark

So would a kitemark or accreditation scheme help improve data
quality?

It is a move that Caroline Worboys, chief executive of marketing
communications business Broadsystem, supports. "As email becomes
increasingly important in the marketing mix, I can foresee a kitemark
being introduced," she says. "The marketing industry is already governed
by much legislation and if we wish to avoid yet more laws, the industry
must regulate itself more stringently. If a kitemark will make
e-marketing practices tighter, this is the direction to take."

Caspian's Hart also backs the idea - as long as the measures that
databases are evaluated against are meaningful, practical and
achievable.

"My worry would be that, for the sake of simplicity, we introduce a
rubber stamp that doesn't really address some of the key failures and it
becomes a meaningless marque," she says.

The US already has a kitemark scheme, called Goodmail, which launched
earlier this year. In essence, Goodmail gets the internet service
providers (ISPs) on board - it currently works with AOL and will roll
out with Yahoo!

in the next couple of months - and then goes out to email service
providers or individual companies and asks them to register with
Goodmail in return for a fee.

It argues that this will give brands an increased chance of getting more
of their email delivered by ISPs such as AOL. Part of this fee is then
paid to the ISP by Goodmail.

Consumers who get their emails through their ISP then see a symbol (in
AOL's case, a ribbon on an envelope) on any Goodmail-certified mail that
arrives in their inbox.

However, such a system is not failsafe. The problem with Goodmail,
according to Ted Wham, senior vice-president of relationship marketing
company Epsilon, is that the accreditation does not take account of the
relevance of the message. If a company pays the Goodmail fee and meets
the requirements, that firm is certified, even if it subsequently
proceeds to send out poorly targeted, unpersonalised emails of little or
no interest to anyone.

"A kitemark will probably help consumers feel confident that when they
get an email that purports to come from their bank, it really is from
their bank," says Wham. "But the kitemark implementations I have seen,
including Goodmail, are not linked tightly enough to the messages. There
is strong evidence that consumers prefer personalised messages based on
their behaviour, whether they have a kitemark or not."

THE BIG ISSUE - How to avoid email blunders

Beat the filters

The most common mistakes relating to email campaigns include subject
lines that read 'insert subject here' or include in the subject line
words such as 'free', 'discount', 'click here', 'offer' and
'opportunity'. These are guaranteed to get the email blocked by the spam
filters of most ISPs.

Email filters examine the words in the subject line and ISPs'
sophisticated anti-spam methods use a points system to identify phrases
and words commonly used by spammers. If an email goes over a certain
number of points, it is filtered out and not delivered to a
customer.

Don't disclose email addresses

The disclosure of other people's email addresses is unlikely to inspire
loyalty among your customers.

In 2002, US-based healthcare provider Eli Lilly accidentally disclosed
the email addresses of almost 700 Prozac users. The company sent regular
email messages to people using Prozac, who had signed up to receive
information, but it hit the wrong button when distributing one of its
Prozac email campaigns, revealing addresses to all other e-mail
recipients in the process.

The company avoided a fine but was told to improve its email marketing
systems.

Get the wording right

It is absolutely imperative to get the wording right in a medium such as
email, where word can spread fast.

In September 2005, Cityboytees.com, a US-based custom T-shirt retailer,
offered anyone who signed up to its email list a full-colour image of
their choice, printed on a T-shirt for just $9.95 plus $3.85 shipping and handling. The offer was aimed at business customers
but this was not specified in the offer. Online bargain hunters took
advantage and the company found itself printing thousands of T-shirts
featuring pictures of pets and children, all for little or no
profit.

To its credit, the company honoured all the orders until the end of the
promotion a month later.

NEED TO KNOW - Questions to ask e-data sellers

1. Data collection: How has the email data been collected? When
consumers registered their email addresses, were they aware they were
giving permission for them to be passed to third-party companies for
marketing purposes?

Was it a double opt-in process, ensuring no-one on the list was in any
doubt that by opting in they would receive third-party marketing
emails?

2. Incentivisation: How were consumers on the database motivated to give
up their email addresses? If they were offered reward points, or the
chance to win a prize, how confident are you that they will genuinely
want to receive information from your company?

3. List performance: How has the list performed when it has been used by
other companies? Can the list supplier provide details of open rates,
clickthrough rates, conversions to sale and return on investment?

4. Frequency: How often are people on the list targeted via email by
different companies? Is there a cap on frequency?

5. Deliverability: What sort of delivery rates does the list enjoy? Is
there a guaranteed minimum?