During evaluation of software products, focus is mostly on product features. In my view, focus should be how implementable features are and how will these get implemented in my bank.

Typical evaluation process goes like this: Three products (believe me! that’s what is humanly possible) are evaluated. Functional and technical features are listed. One may go a step further to allocate weightages. Three vendors respond, each one ends up with more or less similar scores. These scores then get adjusted with commercials and Boom! here is the right product for us.

Now issue with an evaluation like this is, one misses out on two key aspects:

  • Are these features implementable in my set up?
  • Do I need these ‘nice to have’ features (yes yes! in next 5 years)?

To elaborate, a product may have many features but the question is do I have enough facility (read data) to implement the ‘nice to have’ features. Then one wonders, do I have enough facility to implement the basic features! Well, you know, we have the best data warehouse system, our data warehouse implementation has been featured as a success story in vendor’s press release! Ok, all good! Can you trust the data in your data warehouse? That’s when often people lose conviction in their voice. So the alternative is to interface with the transaction systems like core banking, treasury and loan management systems, yes! some data is in Excel as well. Now the next step is a quality check for data in transaction systems. Please read my next blog to understand frequent quality check techniques. I think, this is one of the most important aspect for any implementation- in case of transaction systems, the focus is on past data while in case of analytical systems, the focus is on current data in transaction systems.

The second point is, do I need these functionalities. In my experience mostly ‘nice to have’ features have their epitaph written as soon as the UAT sign off is done. Optimistically speaking not more than 50% of the product features are used.

So now the question is, how do I find right product and ensure implementation success? It is obvious that the commonly used evaluation techniques do not ensure implementation success. We need a change in approach. I think, one of the approaches may be the following:

  • Identify reasonably a product which can provide solution to the problem- can be figured out through one or two demos
  • De-focus from nice to have features- believe me these are distractions
  • Engage constantly with the vendor to understand the implementation process of the core features
  • Be candid about data quality.