±±¾©Èü³µpk10

There is no above-the-line equivalent to sending mail to people who have expressly said they don't want to receive it, or who have moved house or died. Airing a mainstream TV ad at the dead of night on an obscure satellite channel is probably the closest you could get. Both amount to wasted marketing budget - yet while you'd rarely see an above-the-line agency running a brand-building advertisement where there are no viewers, the DM industry still persists in the equivalent mailing practice.

Last year, the results of a consumer survey on direct mail were announced in conjunction with the launch of the new DMA/Planet Ark consumer-facing website www.ControlYourPost.co.uk.

Its results revealed that one in five people reported that they regularly receive mail addressed to a previous occupant and 17 per cent of respondents have received mailings for a deceased family member.

It's little wonder that improving standards in data is a major concern for the DMA - not least if it is to meet the environmental targets set out in the agreement signed with the Department for Environment Food and Rural Affairs in 2002. But it's not just the DMA that should be worrying about data practice across the industry. Any company that uses inaccurate data to communicate with customers or prospects is not only wasting money but damaging its brand in the eyes of the consumer.

The good news is that adopting a more responsible approach to data hygiene should be neither bewilderingly complex nor prohibitively costly. In fact, effective data management should represent a cost saving through tighter targeting.

The first step is collecting the data correctly from the outset. Companies either gather prospect data themselves or through third parties - but either way there's no excuse for accepting poor quality data. Address management systems from the companies such as QAS, Capscan and Hopewiser are the vital first step for ensuring data collected over the phone is accurate and matches Royal Mail's Postcode Address File (PAF).

QAS - one of the biggest companies in this market, and which was recently acquired by Experian for £106 million - also has an internet product.

Called QuickAddress Pro Web, it ensures that only minimal data entry is required to generate a full and accurate address through a website or intranet.

There's no excuse for address data collected over the phone or the web not to be 100 per cent accurate - but the same cannot be said of the customer's name. Stripping out obscene or nonsensical names from a database is vital if a company is to avoid unwanted media attention generated by from sending a mailing to a 'Mrs Fat Cow' or 'Lazy Slob'.

But tracking the public's current favourite spoof names requires constant diligence, according to Chris Duncan, managing director of Alchemetrics.

"These lists change over time. David Brent would not have been flagged up as a potential problem until recently, for example, whereas Loadsamoney is rarely collected now. And pattern-matching is not always an exact science."

Data cleaning is not as simple as it might seem, so it's worth working with a bureau that has dedicated expertise in this area. Duncan adds that, while you may be tempted to suppress a "Mr Daft" from your database (seeing as it's the name of a Mister Man character and is quite likely to be a joke), bear in mind that there is in fact a pocket of around 15,000 people who share that surname in the Nottingham area.

One company that appreciates the benefits of outsourcing its database management is UIA Insurance, a company that offers general insurance products to not-for-profit groups and trade unions. Damian Reilly, head of marketing at UIA, explains that the company has a successful relationship with Bristol-based The Database Group.

"My view is that we are a general insurance company, not a data management company. The skills and expertise that you need are quite specialist and those are skills that are best sourced from out of the business," says Reilly.

The company has a rigorous approach to data quality (see case study, page 10) and it is just this sort of best practice that the DMA is keen to promote. Last year, a working party of industry experts, chaired by DMA chairman David Coupe, was responsible for the launch of The DMA Guide to Suppression. This is available at www.dma.org.uk/environment and includes frequently asked questions, top tips and details of the DMA Code requirements for suppression.

Among the tips is the hopefully reassuring fact that no database will ever be 100 per cent accurate. With 600,000 adults dying each year and 3.25 million people moving home, the speed of data decay is too rapid for any single company to keep up with on a daily basis, but companies should at the very least suppress their database on a quarterly basis.

The solution they choose will depend on the size of the business, but Ann West, sales manager at The REaD Group, suggests that a good starting point for any company is to test the data. "Most bureaux offer clients a free audit that will provide the owner with a good indication of the overall accuracy," she says. "This includes quality of address records, levels of duplication, and percentage match to suppression files. It allows the end user to calculate the cost benefit of suppression."

Higher-volume mailers, West points out, will get the best value from a licence agreement. "Flags are attached to the customer or prospect files, with savings calculated on the frequency of the mailing," she explains.

However, low-volume mailers can benefit from a number of online data cleaning tools now available, including Experian's Intact solution.

Toyota is just one company to take advantage of this service. It has one centralised database for Toyota and Lexus customers and prospects that is cleaned on a quarterly basis through Intact. All negative data - including MPS registrations, goneaways and deceased records - is suppressed and telephone numbers appended if necessary.

According to Mike Sherwen, Toyota marketing database manager, the speed of turnaround enables the company to integrate data cleansing into its mailing schedule with minimum disruption. For any company, minimising disruption is vital if data hygiene is to be successfully incorporated into its direct marketing programme.

But the onus shouldn't just be on the client to make data hygiene a priority.

Everyone in the data chain should take responsibility, according to Hayley Tooke, business development director at list company Wyvern Direct Response.

"We have become acutely aware that the responsibility of suppression and maintaining an up-to-date database lies not only with the end client but also with us," she says. "If a company's portfolio is not in excellent condition, it doesn't take long before word spreads. To combat this problem, many list managers now include contractual clauses insisting that they undertake the cleaning of lists through suppression files as standard."

Using a reputable list broker that has these sorts of contracts in place should ensure that the quality of the rented data is acceptable. Don't hesitate to ask how recently data has been cleaned and for the details of the suppression files used if you are in any doubt. And, of course, no company peddling poor quality data can hide behind it for long. The number of responses, and, conversely, the number of complaints and postal returns, will soon reveal the true value of the file.

Email data is relatively new to the market and quality is still a major issue. The main problems arise at the point of data collection. All too often the amount of space allocated for an email address is insufficient, making it impossible to decipher the correct address. Thomas Adalbert, managing director of The Preference Service, is very familiar with this problem.

The company has been collecting email data through its paper-based surveys for four years and was finding that only 50 to 60 per cent of the email addresses were usable. Now the company encourages respondents to write their email address in capital letters and leaves a box for each letter.

As a result, 80 per cent of the data collected is now correct and the company boasts a 340,000-strong email database.

"Companies shouldn't be scared to use space," comments Adalbert. "By using boxes, people have to separate out the letters which leads to a significant increase in accuracy."

It may sound an obvious point, but companies frequently leave insufficient space on paper-based data collection forms - such as vouchers or competition entry forms.

Getting the basics right at the start of the data process is vital. From then on, there's really no excuse to let customer information degrade, given the number of online and offline solutions now available.

COMMENT

MARTIN SMITH, managing director, Millennium

The use of suppression files should be built into a direct marketing voluntary code because it is the most effective and cheapest way of damping down the junk mail fire. Anything imposed by Westminster or Brussels is likely to be far more restrictive.

RICHARD ROCHE, head of multi-channel retail and media markets, Royal Mail

Consider how to use your chosen suppression files. This could be in-house or through a data bureau managing the process externally. For larger organisations, an internal department to manage the process could be beneficial and cost-effective.

SIMON LAWRENCE, joint managing director, Information Arts

Instead of looking at data quality as an ad hoc procedure when a campaign is about to be mailed out, it should be viewed as a long-term company objective. To achieve this it is of paramount importance that data quality receives board-level support.

CHRIS DUNCAN, managing director, Alchemetrics

Cleaning and suppressing data is the key to ensuring your database remains accurate and consistent. It's often easier for a customer database to be populated correctly in the first place. Many companies capture incorrect customer information at source and fail to rectify this.

TOP TIPS

1. Get to grips with the scale of the data-quality issues you face. Many best practice bureaux will conduct a free data audit for you in order to demonstrate their expertise.

2. Never assume data bought in from a list manager will automatically be clean. The scale of the problem can vary massively - anywhere from a five or six per cent drop up to a staggering 60 per cent.

3. Suppress against a major goneaway suppression file based on non-assumed data - don't rely on any file that uses 'return to sender' information - it's just not reliable.

4. Then screen against a niche suppression file - depending on company- and mailing-specific requirements.

5. Find out how regularly suppression files are updated and what their sources are, so a reliability benchmark can be established. And check how regularly files are updated, since goneaways may just be temporary.

6. Search out bureaux that offer the choice of 'per match' or 'per record'. With per record you pay once for every record on your database, while per match incurs a charge for every duplicate matched. Using multiple suppression files can result in more matches than you have records so can be more cost-effective and predictable to work on a pre-record basis.

7. Find a bureau that can and will run each of the necessary suppression files in ascending cost order.

8. Suppress the data against all mandatory and recommended files, such as TPS, CTPS, FPS, email opt-in, bereavement files and MPS.

9. Suppress returns to your own campaigns. Whether the return is a product of a true goneaway or a disgruntled 'return to sender', those individuals are not going to respond.

10. Plug the gap left by suppressions by using the NCOA, which validates genuine goneaways and then tracks them to their new address.

Source: Steve Clarke, client services director, CDMS.

CASE STUDY - UIA Insurance

UIA Insurance clearly stands out in the homogeneous insurance market.

It provides tailored insurance products for trade unions and companies in the not-for-profit sector and the funds are managed on a socially-responsible basis.

As such, UIA head of marketing Damian Reilly believes that the company has a greater than average duty of care to its members - and that extends to its approach to data quality. "If we were lax with our processes, we really wouldn't be doing ourselves or our members any favours," he says.

Reilly joined the company in 2001, having previously worked at One 2 One (now O2) and Whitbread. This background has helped him implement what he describes as "big company processes" in a medium-sized business.

The company has a database of nearly four million prospects and 170,000 customers and communicates regularly with both groups. Rather than cleansing its database quarterly - which is typical in most organisations - UIA conducts a complete trawl of its database every time it carries out a campaign. This process is outsourced to The Database Group.

"We tend to over-clean and suppress if I'm perfectly honest," says Reilly, "but it really is a benefit, both in terms of cost management and communication effectiveness."

The Database Group uses an extensive suite of suppression products that vary according to the product being promoted. For life insurance products, for example, UIA is particularly diligent. "We clearly wouldn't want to offend anyone, so with our life product we suppress on a household level as well as an individual level."

This recognises that while it would clearly be damaging and upsetting if a mailing about a life insurance product was sent to someone who had died, it would also be insensitive for anyone in the family of a deceased person to receive such a mailing.

This rigorous approach to data quality is clearly paying off. Reilly reports that there has been a "massive decrease" in the number of complaints the company receives about unwanted mail since it has taken on board new data-quality policies. And it is also contributing to commercial success, with UIA enjoying a healthy turnover that is steadily increasing year on year.