Getting your data quality right should be the first step in a structured approach to improving the customer experience. As The REaD Group explains to David Reed in this sponsored feature, it really is a case of getting back what you put in to your data-driven marketing.
Marketing is all about cause and effect. Apply the right stimulus - an offer, creative message or proposition - in the right place and you should get a positive outcome. It is what drives direct marketing and has made it the dominant mode of activity, across online, offline, mobile and broadcast channels.
Yet marketers do not always consider the cause and effect at work in the build-up to running a campaign or customer management programme. By ignoring some of the first principles - especially those that relate to data - they risk failing to get the expected effect. Put the right effort in, however, and those outcomes will improve.
“A lot of people have an understanding about what customer management and loyalty are all about and can see the benefit to the business of understanding who those customers are through insights,” says Glenn Cook, senior consultant with The REaD Group. “So they leap straight into segmentation and use that to drive their marketing activity. But they forget to make sure they have got the right quality data to support that insight and outputs.”
His experience as former head of supporter services and database marketing at Macmillan Cancer Support showed him the importance of pursuing each stage in the right order. “When we brought in a customer insight analyst who understood the market and had experience in the wider world, we saw the charity was making assumptions based on flawed data,” recalls Cook.
Transactional data held on supporters had not been integrated into a single view, so the charity was unable to see when an individual was giving by Direct Debit, but also donating in other ways. “If you are starting to plan your customer journey, you need to look at what others have done. It gives you an indication of where to start, where you can go to and the stages involved,” he says.
The REaD Group has taken its experience of advising, supporting and delivering effective data-driven customer journeys for a wide range of clients and put it into a four-step road map. At each of the key stages - Data, Insight, Customer Management, Loyalty - there are important data issues and actions that need to be taken. It is the linking up of all of these stages to feed back into one another that builds an ever-improving and effective customer experience.
Macmillan’s progress towards delivering a better experience for its supporters began with a focus on data quality. By cleaning, de-duplicating and removing records that had no data in them or were deceaseds and goneaways, it was able to start with a clear view of its actual supporter base.
Taking a close look at the underlying data forces an organisation to think carefully about its data management strategy. “You may not necessarily want to create a single, golden record for each individual - your ‘customer’ could be an individual, or a couple or a family. From a purist point of view, it can be very complicated to drill right down into whether a transaction from an address was his or hers,” notes Cook.
In some cases, couples or family groups will consider their commercial relationship with a brand to be for the whole household - consider a cable or satellite subscription, charity giving or arts organisation membership, for example. The business may need the flexibility to talk to individuals as well as households, which means ensuring that data is held both discretely at individual-level and with links to others in the household.
Often, data will have been added into a system without including variables that allow this dual approach to be adopted. For example, a name might be taken for a customer, but no salutation (Mr, Mrs, Miss, etc). It is also quite rare to note the gender of a customer at the point of data entry, even though this is the best chance to validate that part of their identity, either by getting call centre agents to record it, or by offering onscreen prompts during online registration.
What is surprising is how often organisations do not realise the connection between upstream customer management problems and downstream data quality. “While working for large global oil company on their global retail loyalty programme, it became evident that there were no attempts being made to validate basic contact data such as addresses,” recalls Cook.
“The first sign that an address was invalid was when it was returned as a goneaway. Even then it took up to six months for this to be applied to their core data. No attempt was being made to recover any of the addresses and the records were simply being marked as ‘do-not-contact’. Acquisition of new members was purely passive, relying on advertising at the till in filling stations. The net effect was like trying to fill a bucket with water when there is hole in the side,” he says.
Data enhancements from third parties like The REaD Group can form an important part of this data cleansing and integration exercise, as can the use of its suppression files to remove those deceased records. Improving address quality is one of the core dimensions of this data hygiene stage that has immediate benefits through better targeting and lower wastage of posted communications.
Age is another extremely powerful variable that is often not asked for. Cook points out that including this request has a major impact on customer management. “If you can get date of birth, get it. Avoid age bands because, unless you date stamp when the record was created, they fall out of date. Similarly, if you just ask for somebody’s age, do you know when that was entered into the database?” he asks.
Cleaning and enhancing data is one aspect of the focus on data quality and data management that needs to be considered. It forms an important building block for the other dimension - bringing together transactional and behavioural information for each customer or household.
“For many charities, the typical ask is £5 or £10. Macmillan looked at historical donations from supporters and how they were made, for example whether somebody had given more than £100 in the past three years. Then instead of approaching those supporters for £10, we were able to ask for £25 or even £100,” says Cook.
As with gender and age, care needs to be taken about how this information is integrated. A lot of organisations will have customers who appear to regularly spend or donate several hundreds of pounds. But these individuals could be acting as “agents” for a larger group, raising money from friends and family through sponsored events or coffee mornings, in the case of a charity, or grouping orders from others in order to get a discount from a catalogue.
Building appropriate data sets within a customer database is not limited to demographic, transactional and behavioural items. All of those can and should be captured at first point of contact and regularly added to a file during the lifetime of a relationship. As the customer journey progresses, that history becomes fuller and a better predictor of future activity and propensity.
What they do not provide is an indicator of motivation - the reason why an individual picks one brand or cause over another or why they engage in particular ways and channels. While this can appear complicated to achieve, there are some relatively straightforward ways to build motivational data.
“We created a set of 13 questions that could be answered yes or no via a tick-box questionnaire that took less than a minute to complete,” says Cook. These questions provided an insight into the drivers of giving to the charity, which then created an eight-group segmentation.
At one end of the spectrum, donors were motivated because they had personally suffered with cancer - at the other end, they may just have attended a charity event. By looking at the distribution of supporters across these segments, the charity was able to pick out the leading segments and create personalised messages for these groups.
What this example makes clear is the power that the right data - cleansed and enhanced to the right standard and then appropriately integrated - has in driving customer insight. In turn, those models and segmentations can be used to align the customer management process with the underlying profile, with the result that loyalty increases. That will deliver benefits on the bottom line, which is why Cook notes that, “organisations shouldn’t be frightened to put money into cleaning up their data because it will generate results.”