The Platform Play
There are essentially two different schools of thought when deciding how to approach the applications used within an organisation. One would suggest that you should find the best application for each individual circumstance, that best meets the needs of users and provides the functionality required, with the other proposing investment on a suite of applications that largely meets the needs of users and predominately provides the functionality required. These two approaches are known as a ‘best-of-breed’ or a ‘platform’ play conversely.
It may not surprise that the most optimal outcome for most organisations is a combination of the two.
Differences between each Play
Both approaches have their advantages and disadvantages. A best-of-breed approach ensures that every identifiable requirement is met by the applications, whereas a platform approach ensures that costs and complexities of integration are minimised.
So, what tends to happen in real life?
Do most projects, with a detailed analysis of requirements, confirming that every single possible requirement should be met by a new software package is in fact met? It’s true that a lot of new purchases of systems start with an analysis of requirements. Requirements are often documented and identified as being mandatory, optional or desired. But how often after implementation, are new requirements identified, noting of course, that these new requirements may not have existed at the time of software selection?
Contrast this with a platform approach. In this case, a great deal of planning originally determines that a specific platform should be invested in, with each requirement for an application being assessed against the solution available within that platform. In many cases, the platform product available may not meet every requirement, but being part of a platform or suite of solutions, it is often already integrated to some extent with the rest of the platform.
So what do you do when some requirements aren’t met? There are two options available:
- Give up on the platform solution, go with the best-of-breed product and integrate where possible; or
- Go with the platform solution and develop a custom solution for the missing functionality.
In the past, custom development may have been a recipe for disaster. But with the latest technology platforms embracing custom modification and enhancement, customisations can often be made without substantial cost or concern that changes may not survive upgrades.
Equally, integration can often be easier than it used to be. Modern techniques involve the use of APIs, and specifically “RESTful APIs” which allow for relatively seamless integration between disparate systems. There is always a risk of continual maintenance of integration, but this may be overcome where substantial components of system require best-of-breed solutions to be incorporated into organisational systems.
Importing csv files is NOT integration!
So where does this leave you in search of an optimal solution? The advice is to take the platform play, in so much as it makes sense, with the option of including a small number of best-of-breed solutions where necessary.
Just make sure that even the best-of-breed solutions are integrated too. Having disparate systems that don’t talk to each other leads to information silos and duplicate data entry, neither of which have any place in today’s organisations.