Data Management Tooling

Build, Best of Breed, or Single Tool Suite?

There are many options for organizations looking to support their data management programs. It can be a complex task trying to figure out which way you should go in addressing your tooling needs.  A good strategy for the major components of Data Management should help by identifying current and future gaps in capabilities that need to be addressed.

There are several options – you could try to build your own internally; look for best of breed – i.e., best quality, catalog, governance, lineage tool etc.; or you could look at single suites giving you all capabilities you are looking for. There are pros and cons to all of these.

Unless you have a large technical support organization this one is probably not on the table for you. Not only do you have to build what you need but you also must support it moving forward which is an even bigger task with changes to data platforms, requirements, etc. One other downside is unless data management is your primary business you are taking away resources that could be focused there.

The biggest advantage here is obvious – you best or near-best capability for each component of data management. This method of the last two should come closest to meeting your current or future requirements.  Downsides to this are generally cost, integration requirements, and rate of change.  Do you switch tooling as it does? This change comes at a cost not only in resources but also in constant change for your organization.

The advantages to this are one vendor to deal and negotiate with, similar user experience, and tools come integrated already. The primary downside to this tooling strategy is you get some portions of the suite being very strong with others being less capable – as the old saying goes jack of all trades master of none.  Another possible con is that you give a single vendor more power over your requirements and more negotiating leverage– the more imbedded they become in your organization the harder it may be consider moving elsewhere.

A final hybrid of options two and three is where you see some organizations – a suite for the tooling where it can meet the current and reasonable future requirements and one or more best of breed for particularly complex components like Data Quality or Data Lineage.  You get the pros for the most part of the two options with the con that this will most certainly cost more at the beginning at least.

At different times the strategy of choice seems to swing – for a time best of breed will be very popular and then perhaps when budgets get tight looking at a single suite of tools might become more the norm. Building in house is generally something that is only for large technically savvy shops and in my experience after a time it is abandoned due to the cost in support resources.

In the end your budget, technical resources, and most pressing needs will help decide which of these you go with.  In my experience the biggest mistakes I have seen organizations make is in focusing too much on a single criteria – i.e., Budget, Resources, or Requirements. This can lead to either a system that is cheap enough and requires little integration but doesn’t meet requirements for very long or meet all the possible requirements but must be cut when budget gets tight. Both can require retooling which in and of itself adds costs and resource constraints.

Whatever path is taken, if the tooling decision is based on a sound set of strategies for data management it has a far greater chance of success.