Before you dive in and start building your single view and the structures to support it, there's an important decision to make: what do you want to do with the single view data once you've got it in one place? There are two broad use cases, each requiring a different kind of technology solution.
The first is an analytical single view, the second an operational one. While the solutions required to deliver each are at opposite ends of the Master Data spectrum, it isn't to say you can't start with one and then expand to do the other as well.
Creating a single view is a journey, and once the benefits of your initial work become clear, other areas of your organisation are likely to take note and start wanting to do similar things.
Analytical single view
An analytical single view is primarily a store of your key information, designed to provide insight and improve decision-making in your organisation – we'll look at some of the things it can enable you to do in 'Making the most of your single view of the truth'.
As such, it can generally slot in alongside your existing IT landscape with relatively little initial disruption. Its working will also have little impact on the data storage systems, because the data is usually drawn out in batches; once a day is typically enough to provide sufficiently up-to-date data for most analytical purposes, and these batch processes can run at night when demand on the source data systems is lower.
For these reasons, and because less disruption usually means less upfront cost, the analytical side is typically a good place to start when creating a single view, particularly if you focus on an area that will deliver a quick win and visible ROI.
It can show your organisation what's possible, thereby securing buy-in (and budget) for further, more advanced single view work, including in the operational space.
Operational single view
Any data collected, cleansed and compiled into an Analytical Master Data Management tool will reside there: that is to say, it won't be fed back into the source systems.
If two departments have different information about a customer in their respective lists and this is identified and resolved by the analytical single view solution, the discrepancy will remain in the operational source systems – so those who use them won't enjoy the benefits of the single view.
If what you're after is a way to draw out the data, resolve any anomalies to create a single view, and then feed this back into your source systems so that everything remains in sync across your organisation (as well, potentially, as using it for analytical purposes), then you need an operational single view solution.
By its nature, this is more complex than building the read-only analytical single view, both from a technology perspective, and in terms of Data Governance.
Each source system will likely store the data in a different format and offer different methods of real-time integration.
Tailoring the solution
If they're bespoke systems that were built long ago, do you still have the expertise to make the changes? And if they're off-the-shelf products, do they support what you need to do? Furthermore, will these systems be capable of handling the more frequent load that the Master Data tool puts on them?
Aside from technology challenges, the overall system needs to be set up to deal with situations where a piece of information may be updated differently in two systems: which takes precedent?
Thought will also need to be given to how information is formatted across the organisation, and what terms are used to refer to a piece of data.
It's here you'll need robust Data Governance to ensure the single view doesn't get bogged down with staff constantly having to resolve naming and formatting anomalies, for example.
Get this right, however, and you'll have an incredibly powerful operational tool at your disposal. We'll look at some of the things you'll be able to do with an operational single view in 'Making the most of your single view of the truth'.
Where the operational data resides
An analytical single view solution will typically draw data out of the source systems and consolidate it for use in a Data Warehouse, in such a way that it can be presented to decision-makers.
When it comes to building an operational single view, there's a choice to be made, however. Should the single, golden record be stored in a central database, which all the source and consuming systems connect to? Or should the data remain in the source systems, with the Master Data tool simply pointing all requests for a given piece of information in the right direction?
It doesn't need to be a black-and-white decision. It could be that part of the single view is kept centrally, while other elements remain on their source systems, appropriately signposted behind the scenes.
The decision will depend largely on the source systems: how easy are they to integrate with, read from and write to? Are they be capable of supplying the data required in a timely manner, without affecting performance of their core duties?
- Trevor Hodges is an Information Management Consultant at IPL. He has worked in Information Management consultancy for more than a decade.