Getting Better all the time

Get a customer’s name and address wrong and you will soon hear about it. So why are data quality programmes not universally adopted and strategically important?

For one quarter of you reading this, data quality is foreign territory. With 27.5 per cent of respondents saying their organisation does not have data quality programmes in place, it is no surprise that consumers and businesses continue to be alienated by inaccurate or out-of-date information in the communications they receive.

Yet the changed dynamics of the economy over the last two years have made insisting on data quality investment a more complex challenge. On the one hand, there have been far fewer house moves, reducing the level of basic address change and customer re-linking required. Financial pressures may have led to more relationship break-ups, however – and paradoxically a mini baby boom – which means there will have been more demographic changes.

In the B2B world, the reverse is true – far more changes of address and far fewer births as companies close or downsize their premises, but start-ups falter. Data quality can look like an obvious line of cost to cut when money is short, but it is generally a false economy.

What is interesting about the 72.5 per cent of companies that are running DQ programmes is their reasons for doing so. Top of the list is to gain better marketing performance (named by 77 per cent), which offers an almost instant payback from any money spent on verifying, validating and enhancing data.

Close behind is better customer relationships (75 per cent), indicating that companies now understand how damaging it can be to use the wrong name, address or other variable during contact with customers. Third on the list is better business performance (63 per cent), a more strategic, but also longer-term benefit from better data quallty.These three drivers are the usual basis for any business case about improving data. In hard times, they provide an even stronger motivation and source of return on investment which makes the intangible nature of DQ very visible across the enterprise. A second tier of drivers was led by compliance (53 per cent), better sales performance (43 per cent) and operational efficiency (42 per cent). It is surprising that compliance was not rated more highly as a reason for driving up data quallity, although it probably reflects the more local level at which these programmes are being operated.

Although economies of scale can be achieved, the increase in costs of data quality programmes tends to be linear

In four out of ten cases, DQ is a standalone project being run by an individual function. (Marketing is the lead function in eight out of ten business, with others a long way behind – sales gets involved in 42 per cent of programmes, customer services in 37 per cent and dot.com in only 10 per cent. Support from IT is enjoyed by 53 per cent, but only one in five DQ programmes have the involvement of finance or the board.)
Just over one quarter of organisations (27 per cent) run their programmes at individual department level, but in line with a corporate DQ strategy. Only 17 per cent have a central data governance function that is driving the programme. In contrast, IT is running DQ centrally at just 7 per cent of businesses.

What this picture reflects is that in most cases, data quality is being applied tactically at the point of output – usually when a marketing or customer communications campaign is executed – rather than strategically as part of ongoing business processes. Although this is a good basis for improving data, it can often leave the root causes of errors unchanged.

The vast majority of data quality programmes are focused on improving customer data. So the level of investment being made into DQ will reflect the size of the customer database held. As findings about the single customer view have shown, this tends to be highly polarised depending on whether a company operates in B2C or B2B markets.

The 26 per cent of respondents spending less than £10,000 on data quality are therefore likely to be enhancing relatively small numbers of business records. At the other end of the scale, 5 per cent spend more than £1 million each year. Although economies of scale can be achieved, the increase in costs of data quality programmes tends to be linear with the number of records being managed.

If there is one area in which the gap between tactical applications and strategic goals can be most clearly identified, then it is around how well organisations measure the financial benefits of their DQ initiatives. The expection might be that ROI is obvious and easy to measure, yet the reality is different.

Asked to rate their company on this, 44 per cent were neutral, with only 26 per cent being positive, compared to 30 per cent rating their measurement negatively. The overall rating was 2.88 out of 5 – below average. Difficulties in associating costs with poor quality data – and then measuring the improvements once quality goes up – are part of the explanation for this poor performance.

Certainly companies do not lack the tools to do the job – address management software and CRM systems are present in two thirds of business. Less use is being made of online or bureau-based data cleansing services. Most significantly, only 9 per cent have data quality monitoring software in place, which makes that ROI measurement easier.

What is clear is that data quality matters and is recognised as important to the organisation – 35 per cent say it is very important and 35 per cent quite important. Any business that has direct contact with customers can not fail to recognise this fact. Even so, there are still challenges to be resolved, especially as half of companies rate their satisfaction with DQ tools as only moderate (with an overall rating of 3.44 out of 5, or just above average). As with the need to drive up adoption of DQ, among those already running programmes, there is still work to do.

Those companies that are not running data quality programmes are not blessed with perfect data. Nearly all of them (96 per cent) acknowledge that marketing might be impacted by poor quality data, while 83 per cent see its impact on customer service, 78 per cent on sales and 65 per cent of customer insight. Four out of ten also recognise that their Dot.com operation and finance department could also suffer problems.

The most likely damage is to customer relationships, followed by higher marketing costs.

The most likely impact is damage to customer relationships, which 87 per cent recognise as a risk. Just behind this, 78 per cent named higher marketing costs and 74 per cent potential loss of sales. A more intangible risk was also identified by 65 per cent in the form of potential damage to brand reputation. The same proportion were also concerned about inaccurate business reporting, which exceeded compliance failures as a worry. Only 39 per cent see any possible link between inaccurate data and fraud.

Where an impact is noticed depends on where the organisation is looking. For companies that are not running a DQ programme, email bouncebacks are the main indicator of poor quality data (87 per cent). Six out of ten see it in postal returns, unavailable telephone numbers and customer complaints. Surprisingly, just 17 per cent mentioned unpaid bills and 9 per cent said there were no indicators.

Awareness of the tools to improve data quality is high. In fact, most of the companies not currently tackling data quality expect to do so soon – 32 per cent within the next six months, 41 per cent within a year and 18 per cent within two years. Even companies that are confident in their data can not ignore DQ for long, it seems.

As with those already tackling data quality, the same key drivers are named by non-users as a reason for laying plans in the future – better marketing performance (77 per cent), better business performance (54.5 per cent) and better customer relationships (41 per cent).

Cost is the main obstacle for earlier adoption of a DQ strategy, according to 41 per cent, although 36 per cent named a more strategic problem – the lack of internal process. Money is always an issue that data managers have to confront. If the economy turns, then this will become easier.

Getting top-level buy-in to change how data enters the business is a much tougher problem.