Data-savvy businesses often worry about two things. First, they worry about data quality. With data so integral to so many parts of the business, a lack of quality control can have massive financial consequences. Second, they worry about where the data is. Their data may be perfectly correct, but different in different places – and nowadays, there is so much information across so many systems that it becomes hard to bring it all together meaningfully. For example, my name is legally Jacob D. Freivald. If I’m signing something off officially, I would write J. D. Freivald. If I’m addressed by a friend, I’m always Jake (sometimes I’m listed as Jake in CRM applications, too.) Add in nicknames, social media handles, and so on, and a company can be confronted with numerous personas that are all, in fact, one consumer.
As a consumer, of course, I expect to be treated as one customer: for example, whether I present myself as Jake or as Jacob, I should receive the same offers and discounts. Companies who have me for a client need to correlate information across their different systems to get a 360-degree view of me, regardless of differences in names between applications.
At the same time, they have to make sure that they don’t correlate my data with someone else. Family members who are James (or Jim), Jon (or John or Jonathan), Joseph (or Joe), along with the possibilities of Juniors and Seniors – maybe living now, or previously, at the same address – can really throw a wrench in the works.
This is why businesses depend on a “master record” of what the system knows, populated with all this information, and governed by data stewards. In the e-commerce, omni-channel era, this kind of attention to detail – master data management, or MDM – is essential for guaranteeing a seamless consumer experience, free of irritating snags in the customer journey. So it is crucial that these different accounts are consolidated.
Note that this is different from other uses of data warehouses and similar analytical data stores. Data scientists and analysts benefit from MDM because they get a more integrated view of the enterprise, but they’re not the ones who drive operational value. Business people with operational responsibility need to be involved in determining what is and isn’t important.
MDM can be a daunting process, and it is often seen as long and complex, especially for small and medium businesses (SMBs) which often have a relatively small IT department.
With that in mind, here are a few top tips to ensure that your efforts to install MDM aren’t wasted:
1. Find a business champion
It can sometimes be difficult to get the support you need for an MDM project, especially when budgets are tight. It is therefore important to make sure that you have someone to turn to on the executive team who actually cares about the data and values seeing it used effectively.
This “business champion” is essential for ensuring that your project is carried through to the end, because they will back you up at a high level and ensure you get the support you need. You need someone to be singing MDM’s praises in the boardroom.
2. Enlist a data steward
Once you have your c-level executive on board to sell your case at a higher level, you need to make sure you have someone implementing change at the next level down. This “data steward” should be from the operational side of the business – so not a data scientist – but they act as a go-between for the business and the data.
This allows the data scientists to get on with what they do best – mining for data – and not have to worry about making decisions about specific records, field values, and individuals. It’s crucial that the steward takes responsibility for making those decisions and follows through with all areas of the business.
3. Model clearly
Many enterprises begin their data management project with a desire to have a single view of the customer. With this in mind, it can be tempting to dive straight into the data without establishing goals first, but there is no value in exploring if you don’t have the end in sight. To get value, you need to focus on what you want to achieve at the end of it.
Modelling the so-called “golden record” – the view of the business entity that you really need – is important with MDM projects as it allows the business to choose which direction they want to go in. Of course, it will always be a bit of trial and error, but it is easy to go back and change the model, and changes will happen more quickly if the modelling is thorough. If it begins with the business needs, then it’s guaranteed to provide better business results when deployed.
Jake Freivald, vice president of marketing at Information Builders
Image Credit: Shuttterstock/Bruce Rolff