He then discussed current macro and micro-economic trends (we’ll talk about this more in a second) and how organisations could look at a DQ focused Single Customer View (SCV) to provide greater time to value than its more typical sponsor, MDM.
Traditionally, organisations have tackled their SCV requirement through the deployment of an MDM platform. And yet, as Philip discusses in his paper, ‘MDM has always been complex, costly and time-consuming to implement’ and so not necessarily, therefore, in tune with modern business requirements. Layer in an increase in regulation and we have a perfect storm of reasons for organisations to seek an alternative route.
So, what options are there for organisations looking to keep costs to a minimum or take a more agile approach to developing a SCV?
Data quality can sit at the heart of your SCV
Experian has long preached the benefits of data accuracy. Accurate data = accurate decisions = multiple tangible and intangible organisational benefits. And yet, many organisations are still unaware of the wider implications of well-managed, fit-for-purpose business data. In his paper, Philip highlights that good quality data can save you money, make you money, support compliance efforts, and support a wide variety of IT and business processes. That sounds like a pretty good list to me and shows us that DQ should be taken seriously at an organisational level.
Philip showcases the three stages required to build a SCV.
- Identification of data sources and systems;
- Assessment of data accuracy levels and implementation of transformations (matching and profiling to link and de-dupe data);
- Monitoring accuracy levels over time to ensure data remains fit for purpose.
Encouragingly, all three of the above activities can be completed using data profiling and data quality technology. And given the nature of this type of technology, integration periods can be dramatically trimmed vs. traditional MDM implementations.
Where do I start on my SCV journey?
At Experian, we approach building a data quality-focussed SCV in four phases. I won’t go in to each one here today but you can find more information and resources on our single customer view page. Today I wanted to focus on the questions you should be asking yourself and your business as you build out your SCV framework:
Phase 1 – Investigate your data assets
- What data sources do we want to pull into our SCV?
- How complete/trustworthy is each record?
- Do we have duplicates?
Phase 2 – Assess accuracy and impact
- How accurate is my source data? (Don’t know? Then you need a profiling tool)
- What is the knock-on effect of not fixing each issue found?
- Which attributes will be useful in the matching process as match keys?
Phase 3 – Improve your data landscape
- What technology will I use to cleanse, match and transform my data?
- How will I validate my data against referential data sets? (we can help)
- How can I manage large volumes of data?
Phase 4 – Control your on-going data quality
- What measures will we have in place to ensure our data stays accurate and free from duplicates?
- What triggers will we have to tell us when these thresholds haven’t been met?
- How will we educate our staff on the value of accurate data?
There is a lot to do before May 2018
Building a SCV gives you an opportunity to serve clients efficiently and accurately to help build trust and loyalty. Getting a true sense of your data will help decision flows within your organisation which has been shown to positively affect staff morale.
And, with the GDPR clearly on the horizon (deadline for compliance is May 2018) the Bloor paper points out that given the timescales involved, implementing a hub-based MDM solution will simply not be practical compared with adopting a SCV then implementing a single GDPR solution against it. This can play a large part in showing the ICO that you value the personal data that you hold and are using it responsibly.
Given the current myriad of internal and external pressures, organisations need a timely, robust and secure solution to managing personal data. And, as Philip Howard says, “What is the fastest route to value? We [Bloor] would suggest that it is in focusing on data quality in the first instance”.