Data, and how to work with it

Do I need accurate relationship discovery in my project?

Written by Phil Watt | 15-Mar-2024 02:09:28

Our partners at DataJPS often ask us: how do we know if Relait or Kindred would be a good fit for my customer? ‘Open and transparent’ is one of our values at DataJPS and, in that spirit, I’m sharing here what I tell those partners.

The table below shows the question, the rationale behind it, and an answer range that indicates your candidacy:

Question Rationale Preferred Answer Range
Do you have a data intensive project with multiple datasets? This could be any one (or combination of) data migration; digital transformation; analytics; AI; Customer Data Platform; data warehouse; etc

If you do, you are about to feel the pain caused by incomplete information about the relationships in your. Getting an accurate picture of the relationships in your data will reduce the pain and risk.

This is the only question where a ‘no’ answer probably rules you out

Yes
How many people work in the back office? Once the number of people supporting the field organisation grows above Dunbar’s Number, it becomes very difficult to maintain relationships with them all. Knowledge management becomes increasingly difficult More than 150
How mature - and reliable - are your data governance capabilities? Relait requires little or no existing data governance / management infrastructure in order to have an immediate high impact. As data governance maturity grows, decision makers may be satisfied with current capability (data lineage, metadata management, data catalogue, etc.). How confident are you that your data governance systems support business and risk management goals? Low / Medium / High
How successful have you been with recent (previous 24 months) data intensive projects? (Time to Value, Budget, Benefits Realisation) If you haven’t felt the pain (or don’t remember it), it is harder to persuade you that the pain is coming, but you can avoid it Low / Medium / High
How many data sources (databases / applications) are you working with? Identifying relationships within a single database is hard. As you add data sources, the problem grows exponentially More than 1
At what level is the application landscape mapped and interdependencies understood? One of Relait’s strongest use cases is in medium to long term planning, particularly around enterprise architecture. By scheduling regular data scans through Relait, you can build up a time-series analysis of how interdependencies are shifting over time, examine trends, and extrapolate future scenarios, to give you confidence in how you can manage change in the application landscape Low / Medium / High

And that’s about it. Probably only the first one rules you out if you say no, because you typically need a project to attach investment against. Even then there is value in making a capability investment in your architecture.

If you want to know more about how we can help accelerate and de-risk your projects, get in touch using the form below.