What is a MultiValue Database?

MultiValue database is a type of NoSQL and multidimensional database, often referred to as a Pick database – eponymously named after one of the lead developers: Richard A. (Dick) Pick. MultiValue databases are also known as NF2, non-first normal form systems.

In defining MultiValue database systems, it’s important to state what they aren’t: a conventional relational database. And while MultiValue systems share many similarities to relational databases, their schemas are less rigid. That means MultiValue databases can assign more than one value to a single attribute.

MultiValue databases are often categorised with MUMPS within the category of post-relational databases, despite the data model pre-dating the relational model. But unlike SQL-DBMS tools, most MultiValue databases can operate with or without SQL.

What are the Advantages of MultiValue Databases? 

The tabular structure of MultiValue databases allows multiple values for attributes whereas relational databases store one fact within a table cell. Everything possible in a relational database is possible in MultiValue, but with the added advantage of storing multi-dimensional data. The multiple value advantage of MultiValue is one of many reasons why they are a popular choice with large data sets typical in large enterprise and government.

Big data analytics tend to run faster and easier on MultiValue systems due to flexibility when dealing with complex data structures. MultiValue databases can manage certain data structures that conventional relational databases find complex. Decades of big data commonly found within large financial institutions and government agencies are why MultiValue databases offer inexpensive data solutions compared to relational database solutions.

MultiValue versus Relational Model 

The following diagram is a hospital’s application for controlling their ambulances. When modelling the call logging and dispatch elements using the relational model, the process equates to nine database tables with fifteen interdependencies.

“With several attempts at data governance for our business, nothing seems to stick.” 

Data governance has become more important than ever. The amount of data being used in business is increasing exponentially. With data being used to drive critical business processes, governance still remains a challenge for many companies.

So why is data governance so hard to get right on the first attempt? 

As the ARQ Group’s Principal Consultant for Data and Analytics, I have seen many times data management teams spend months – if not years – writing slick policies and procedures covering every use case and scenario before trying to apply those policies to the business. 

The focus on every scenario can send data management teams down a rabbit hole, where timelines creep up on the team. This results in change management and embedding activities becoming squeezed. Many of the key players in the financial industry fell into this trap when regulatory pressure forced them to start data governance programs without fully tailoring said programs to the culture of the company or desired outcomes.

Rather than focus on the theory of data governance, perhaps focus on the application of data governance. Remember what you are implementing are rules and regulations – and like all rules and regulations – unless there’s clarity on the WHY, and your people accept and agree, pushback may ensue. This can sometimes amount to frustration, and worst case: rebellion (or at least non-compliance). 

Your WHY is key, but also needs to be balanced by WHAT you are implementing. By WHAT, the level of data governance (rules and regulations ) you’ll be implementing. If you’re implementing a complex set of stringent data governance policies, your WHY needs to be rock solid! 

Alternatively if your approach is to progressively implement a set of simple guidelines, your WHY can also be simple, and gain strength as the business enjoys positive results from your guidelines. 

Never has this balance been more evident as seen in the world’s handling of COVID 19. An initial strict lockdown in Australia was accepted because the WHY was so compelling – lockdown or face COVID19 spread like wildfire with deaths imminent. 

Would Australia have accepted a lockdown if the stakes were not so high? 

The initial WHY is so important, because it’s the point you’re asking people to comply with only the promise of favourable outcomes. As time progresses, if your data governance rollout is successful, your people will see evidence in the benefits of compliance, and likely to follow your policies and procedures. 

Again, we saw this behaviour in COVID19 – by December 2020, an entire lockdown of New South Wale’s Sydney Northern Beaches was accepted. The WHY was accepted during a peak social time (Christmas and New Years), because the results of non-compliance were evident in other countries who had not controlled their outbreaks (as well as Australia).

Getting back to your WHY. How are you going to convince all levels of your organisation that data governance needs to be implemented? 

Your first step is to align with the leaders of your organisation and understand why data governance is important to them. Sometimes this can be easy, where external factors such as regulatory pressure, are top of mind thoughts for the leadership team. At other times you will need to go through a number of possible benefits/drivers to understand what resonates. Benefits/Drivers could be:

• Productivity – The business has a large workforce of people using data, often inefficiently and in a duplicative manner. Data governance can help streamline and ensure data is used and processed once for multiple uses, reducing overall costs
• Alignment/Agreement – Different areas of the business may calculate key business metrics differently, which in turn could drive different business actions. Data governance can ensure that a CEO does not receive two conflicting versions of key information, and defines who is accountable in the business for that information
• Value – Data may be underutilised in the business, perhaps by the accessibility or quality of data, and your business may not be extracting as much value as it could of, out of data driven initiatives
• Regulatory/Audit as described previously

Once you’ve agreed the WHY with your senior leaders, be clear on the funding and sponsorship they are willing to provide to implement. Resonating is one thing, putting up the investment needed is another. 

Data Governance is not easy and requires hours of work, either from new resources the business will need to fund, or by adding work to resources who are already performing a full time job. When companies refer to failed data governance rollouts, many times this is the reason – companies spend the time to write the data governance policies and procedures, but don’t invest in the capacity for the people on the ground to follow the procedures of data governance, and the whole thing falls apart.

So now you have the WHY and the sponsorship, you can start to shape the WHAT.

Here the key is tailoring the WHAT to your companies specific WHY and funding (hint: here’s where a consulting company like ARQ can help). 

Having been through this process multiple times, we know that if your key WHY is productivity and making data more accessible, a compliance heavy program centred around Collibra and requiring all data decisions to be made by a central data governance office is probably not the model that will engage your business. 

Conversely if the key WHY is being compliant with your regulators, a light touch governance process, focussing more on agility may not be the best model. Throw in other key factors such as the culture of your business (i.e are they a command based culture, where edicts from the top are largely followed, or is it more of a collaborative culture where employees are encouraged to innovate) and it is clear that tailoring a data governance solution for your company is the only way to go. 

Our work at the Australian Broadcasting Corporation (ABC) has started on this journey and very cognisant of the above factors, which we believe will lead to a successful rollout over the coming months and years.

If you want to talk data governance and not sure where to start, the ARQ Group can help.

I’m Lawrence Eva, Principal Consultant of Data & Analytics at ARQ Group. You can connect with me on LinkedIn and I’ll show you how to get started on a sensible, simply and low-risk approach to your data governance. 

Alternatively, visit our contact page and leave your details. We’ll get back to you with the quickness.