Hi community,
I work as the Customer Insights Manager at a Retailer company with 1000+ employees.
I'm just starting to implement a Customer 360 view but I have customer data spread across multiple data platforms.
What elements should I include and which I shouldn't? Please advise.
I'm trying to keep it as simple as possible without having to go back later and re-invent or add.
Hi @Mark Hunt,
A problem statement that you have provided is very common while building the consolidated Customer data view.
All projects start with the requirements. In the case of Customer 360 initiatives, the requirements should cover at least the following aspects:
- Logical data model of the customer domain
- Identification of all entry points of the customer data
- Assessment of which pattern your MDM system will follow
- Identification of the data owner of the customer data domain and all other key stakeholders: process owner, data architect, data stewards and many others (depending on the size of your company)
- Physical models of all systems that will be used to integrate customer data into your Customer hub and all systems which may consume customer data. Ideally, you should think of exposing the data via generic APIs, so that later you can add additional downstream systems with little effort.
- Collect all policies, procedures, standards (including cleansing, enrichment, standardization, etc) to govern customer master data
- Collect all match & merge rules to build the golden records of the customer data while integrating the data from multiple sources.
I am sure there's much more. Just ask if you need further details.
Hi @AnilKumar21, @InitZero, @Monirul Islam, @Jaime A. Rendón and @Harisha Acharya,
Can you please chime in here and share your knowledge with the community? Thanks.
Hi Mark,
I wish you're well and safe.
Before you start, you should use a very good data prep tool.
For that, in terms of elements, you should analyze 2 things: (1) which elements are common for all platforms and (2) which elements the business requires as mandatory; from this point, you will be ready to prepare a good model to move forwards.
Later, you can add more attributes for advanced analytics, but start with mandatory elements.
Good luck!