Why is MDM quicker to implement on CluedIn, in Microsoft Azure?

There is a reason why "time to value" is such a huge advantage when implementing technology. Time-to-value brings benefits such as

• ?Less time consuming cloud costs on something that is not live and actually providing value.

It is not a secret that CluedIn is 100% focused on being the most native Master Data Management solution on Microsoft Azure. In addition, it is not surprising that as part of this involves being the quickest solution to implement. At CluedIn, we have a key ethos we use to build CluedIn that has served us, our partners and our customers for sometime, which is to not focus on making parts of the MDM process simpler or better, but making it obsolete. This ethos has led to innovation such as the popular "zero-modelling" in CluedIn. We think this ethos is the key to making MDM much quicker to implement, as you could imagine, what is the fastest way to complete a task? It is to not even need to do it in the first place.?

This is not an odd ethos, as there are numerous examples of this in technology already. For example, if you had the choice between an intuitive and fast backup service for your laptop, or you had the choice of it being automated on your behalf - I can't imagine anyone would want to previous option.?When was the last time you manually backed-up your laptop?

The first reason for MDM being much faster to implement on CluedIn is at the core foundation. CluedIn is a cloud-native solution. Yes, I know everyone says this.

But in a space like MDM where the average vendor is 14 years old, it becomes pretty much close to impossible for any solution on the Gartner Magic Quadrant to be built and designed for the cloud.

This cloud-nativeness simply means, it is easier and quicker to install and get started. In addition, it is also easier to fit into the expectations of the cloud in regards to pricing, scalability and more.

Installation is only the first step, but in the MDM space, traditionally this process is drawn out, long, time consuming and expensive.?

So then ask yourself, what is the first thing that happens after you have installed your MDM technology? Well, typically this where you start modelling your MDM model. Now, make sure you get this right, because changing this later in MDM will probably be improbable, if not economically impossible. This is the first oddity of traditional MDM that CluedIn has not made faster or simpler, it has removed the need to model data upfront like this. Not only is it theoretically impossible to actually know what this model will look like upfront before we start ingesting and analysing data, but it is time consuming - to the point where it is a well known thing in MDM that "if you get your first model done in the first year, then this is good". There is not one shred of that last statement that sounds like it yields "quick time to value".?

The next big time-drain in implementing MDM is to identify problems in your data first. There are many ways to do this, some more sophisticated than others. One of the tried and true ways is in the form of business rules. So, you have just installed your MDM technology, you have just skipped the modelling due to CluedIn and now you are asked to build up-front business rules as to identify, if not, automate the fixing of bad quality data. You have just run into your next time-consuming process of implementing MDM where you are asked to "come up with all the business rules you need to monitor." Like anyone, you would start to look at your actual data, analyse it and try to spot issues or at least put in some possible errors that you would like to catch such as empty values, null values and more. Once again, this doesn't sound like something that will happen quickly, or yield a good result. After implementing MDM at over 40 different companies myself, I can confidently say that you don't know what you don't know. Yes, capturing common data quality issues is great, but is in general 5% of the actual problems that you will see.

The biggest problems you will have when you bring data together from multiple different sources is "denormalisation". This is why CluedIn accelerates this entire process by analysing the data and offering up suggested business rules on your behalf. This yields two main values. Firstly, you are not the one that is required to find the data quality issues. Secondly, it means that once you fix a particular problem once, you will never need to fix it ever again.?We also can't forget that in the process of having these business rules automatically created for you, you are also cleaning the historical data in the meantime as well.?

This sounds like an odd time to already start talking about sharing your master data with your business, but it is the "key" to acceleration. CluedIn is designed as a streaming platform. This means, it has an ethos that it is "ok" to let data out from your MDM system that is not perfect. This is a controversial point as I am sure most companies buy MDM because they don't want bad data to get out into the operational parts of the business. In saying this, it is impossible to achieve this, even with an MDM system in place. Why?

Because there really is no way to determine if data is "right" or "wrong". If there was some easy way to write code that can guarantee if an address is correct or not, then we would not be in the position where MDM even exists. Now that we have established that it is effectively impossible to achieve the concept of perfect data before we share it, then we can now move on to the expectation and realisation that sending data out into the business without having it perfect is what we should be focusing on. In saying this, if you know what is considered bad data, then you can easily setup rules in CluedIn to block that data going out if needed. Sharing data from CluedIn to other systems is dead easy.?CluedIn uses the idea of "we will talk to you, not the other way around". This means that other parts of your business are not required to even know CluedIn exists. With this "push" model, MDM is accelerated due to the fact that there is no need to introduce big change. If your teams are comfortable operating in data tools, instead of asking them to get their data from a new place (like CluedIn), we can have them continue to work in the tools they are comfortable and we will send the data to them.

Part of implementing MDM is about building relationships between records. The challenge with this traditionally has been that you are tasked with establishing the relationships. This is yet another reason why MDM takes so long to traditionally implement. This task of building a web of relationships is fraught with complexity, especially in MDM systems based off structured and schema-based modelling. So, in true CluedIn style, instead of making the modelling of relationships faster or easier, we don't require or ask you to tell the system how records relate to each other, but rather we automate this part for you. This is important for many reasons. Firstly, it is because in most cases you will model yourself into a corner if you start to hand-stitch data together. Once again, I can say this after implementing over 40 different MDM solutions - evolving a data model to cater for relationships in evolving datasets is impossible to achieve with any level of economic feasibility. Why? Because records rarely relate to each other in the same way across different data sources. Even if it was to, things change. Evolving a schema over time is extremely hard to achieve in a way that also yields quick time to value.?

The final point of acceleration to touch on is Microsoft-Azure-specific. As of CluedIn version 3.3, CluedIn integrates natively with 27+ Azure services. This native integration takes the story of MDM and elevates it to an entire data foundation, an ecosystem of services that weaves together to accelerate innovation. These services range from expected integrations such as authentication and authorization with Azure Active Directory, to native flows of data to downstream systems like Azure Synapse, Power BI, Azure ML and more. The main acceleration comes from the ability to hook into an ecosystem of tooling to solve a problem, instead of just relying on a specific technology to solve your problem.?

The other reason why CluedIn is quicker to implement is because it is has done to MDM what Agile methodologies did to project management. It brought the concept of not having everything in your MDM initiate to work in a big bang concept, but rather implement and gain value from small and incremental wins.?

This agility can only come from technology that can embrace this style of agility. With this in mind, one of the accelerators that CluedIn brings to the MDM space is the fact that we don't expect you to have perfect data in the first place. I know this sounds odd, but many technology solutions will require you to have your data fit perfectly into your newly defined model to even enter the MDM system in the first place. MDM systems should not only be seen as the place where your good quality data resides, but also a safe space for data, where it can land and be given the attention it needs to raise to a level of quality that is needed. This also means in CluedIn that we embrace the agility that data does not need to be perfect before we start using it!?

Traditional MDM is very waterfall in nature, in that it asks you to have your data model finished first, and you can't start the other parts before that is done. All of the above reasons and more, is why implementing your Master Data Management initiatives on CluedIn and Microsoft Azure is a great choice for building an infrastructure of high quality data proliferating throughout your business and allowing insights to be driven off mastered, governed and trustworthy data.