Building a single view of the customer by integrating multiple data sets is still a minority pursuit – but only just. According to respondents in the survey, 57 per cent do not yet have a SCV in place, compared to 43 per cent who do. Given the complexity of creating this strategic data asset, it is no surprise that the majority of companies have yet to create one. More surprising is the level who have.
One of the chief explanations for this high proportion is likely to be the size of companies involved in the survey. Asked how many records are on their SCV, 32 per cent said more than five million, while 22 per cent stated between one and five million. In other words, they are relatively large organisations probably operating in the B2C space. Finding budget for a SCV project is undoubtedly easier in this type of business.
At the other end of the spectrum, 22 per cent of those companies with a SCV have fewer than 10,000 records on it. A further 10 per cent are holding between 10,000 and 100,000 records. These are likely to be B2B companies where the value of an individual customer can be very high, making cost-benefit justifications easier. It is in the middle ground where the scale of business makes investing in customer data integration harder.
For those companies which have built a SCV, support for the project appears to be strong. Three quarters of respondents said that the single view was extremely important to their organisation, with 21 per cent saying it was quite important. A bare handful were neutral about the database and none were negative. With an overall rating of 4.74 out of a possible 5, this makes the single view a very positive thing to create.
Where a weakness can be detected, however, is around satisfaction with the quality of information that is held on the SCV. Although 21 per cent described themselves as extremely satisfied, the majority (55 per cent) were only quite satisfied and a sizeable group (18 per cent) were neutral. A handful were even not very or not at all satisfied. The overall rating for data quality on the SCV was 3.89 out of 5 – definitely positive, but not overwhelmingly so.
Any issues about the accuracy or completeness of data do not seem to stem from the frequency of update. The rate at which information on the SCV gets updated varies by company, but also according to the source of the data – 37 per cent of those with a single view said this dictated update frequency. However, 29 per cent are getting real-time loads and 5 per cent near real-time changes. Daily feeds are in place for 10 per cent and weekly for 15 per cent of data marts, so there are few process issues to explainany dissatisfaction.
A more likely reason for problems lies in the upstream sources that feed the SCV and possible errors or broken processes. The main source of information is extracts from operating systems in 66 per cent of cases, with nearly half (49 per cent) also taking an extract from sales records and 46 per cent extracts from customer services records.
In all of these sources it is possible for the emphasis to be more on getting a business process completed quickly than on capturing customer data accurately. Significantly, only one in three SCVs are being enhanced with third party commercial data (34 per cent), online customer survey data (29 per cent) or extracts from financial records (27 per cent). This is likely to limit the depth of information and therefore the satisfaction of users.
Difficulties with too many data owners, mergers, poor processes and data not being captured at customer level.
It is notable how well populated email address fields are, with 93 per cent of SCVs holding this variable. Since only 63 per cent have a home address, 41 per cent a trading address and 34 per cent a delivery address, a lot of emphasis is being put on digital contact, rather than geographical location.
In a sign of strategic value being driven out of the SCV, 78 per cent hold purchase history by product and also contact history, while 63 per cent have purchase history by value. Yet only 51 per cent have applied a segment code, one of the more obvious areas for a value adding programme to be carried out.
This may explain why marketing planning and campaigns and customer management are the three dominant uses of the SCV. By contrast, only 33 per cent are using it for business planning, 25 per cent for demand forecasting and 13 per cent for risk management. Even so, nearly half (47 per cent) of those companies with a SCV intend to invest more in it this year, while 37 per cent will continue current levels of expenditure. Just 3 per cent plan cuts in funding for this asset.
So what of those companies that have not integrated their customer value into a single view? It is not for lack of customer data – marketing holds a standalone customer database in 66 per cent of companies without a SCV, as do 45 per cent of sales and finance functions and 36 per cent of customer services operations. Even 13 per cent of dot.coms have customer data.
While it might be assumed that this is simply a spreadsheet of customer data, this is only the case at 38 per cent of companies. Although a quarter have either an aggregated report on customers or an ad-hoc file, 55 per cent of those who have not integrated their customer data fully are using an externally-built database.
Obstacles that might also be assumed to exist, such as cost, skills or culture, are also not as dominant. While 23 per cent said they had no specific budget for a SCV build and 21 per cent said it would be too expensive, only 19 per cent mentioned a lack of dedicated resource or company cutlure as an obstacle.
Instead, the barriers to SCV tended to be quite diverse with 34 per cent giving other reasons. These ranged from difficulties with too many data owners or organisational mergers through to poor processes and data not being captured at customer level.
Nevertheless, there is a recognition that creating an integrated customer data asset might become strategically important. Six out of ten companies said the need to improve customer experience might drive a SCV project, with customer insight and improving marketing performance named by half. Four out of ten said they would gain competitive advantage and three out of ten named regulation and compliance.
To some extent, those companies yet to build a SCV are looking at their peers who already have the asset in place to see whether it was worth the effort. While it can not be claimed to be a trouble-free blessing, the evidence of this survey is that the benefits do outweigh the challenges, even if integrating customer data does often expose a further set of problems deeper in the organisation.
Acxiom’s 4 steps to a Single Customer View
A well implemented SCV provides the “truth” about your relationship with a customer, and the tools to analyse this. To do this, it needs to be:
- complete – all the customers need to be included
- accurate – clean and validated data that is up to date
- comprehensive – a broad view of all transactions and interactions accross all channels
For organisations with multiple product lines and channels, building an SCV is a major project, and as a result, many organisations choose to build in phases, and many SCVs are still a work in progress.
Based on our experience, Acxiom would recommend 4 steps for building an SCV:
Step 1: Data Audit
An initial assessment of data quality should evaluate the source system data for completeness and accuracy. This will confirm the data can be used for matching at a customer level and identify any specific cleaning requirements.
Step 2: Clean and Standardise data
Armed with the knowledge from the Data Audit, you can configure the data quality tools to standardise the customer data. This will include: standardising the address format to match PAF and adding or correcting address elements such as postcodes. It can also include checking dates, coded fields, values etc. and rejecting data that are outside agreed ranges of values.
Step 3: Match and Integrate
The cleaned records can then be matched and a summary customer record produced. Acxiom has a knowledge based approach to matching that uses our consumer reference database, to give a more accurate match than can be achieved by algorithms alone. Once records have been matched, you can then create the summary record for the customer, aggregating the available data.
Step 4: Monitor
Customer data is dynamic and changes regularly. Systems also change, as new products are introduced or new systems added to support the organisation. To ensure the quality of the Single Customer View the data quality needs to be monitored on a regular basis.
A Single Customer View can be used for many purposes:
- understanding customer channel preference
- identify customer buying multiple products and target similar customers for cross sell campaigns
- credit risk analysis
- customer insight and analysis