The next characteristic around technology we want to discuss is how much data should you convert? The problem is everyone wants all of it brought over, which can end up being one of the largest risks in a project if it can't be properly controlled. Although you may only be asking for 10 years of data out of 40, the reality is that it's incredibly complicated to bring all of that over.
This is why we have to sit down and have a very honest conversation with our client about what data they truly need. Keep in mind that we always have the option of putting that data in another database and accessing it only when needed. Some of this is about making the team feel comfortable with everything and understand that you're not going to lose this data forever. You'll still be able to access it, it just might not be as clean. Is it really worth that effort to be accessing that data from 10 years ago twice a year?
Something else to consider in terms of data quality is that if you convert a decade of data from a legacy system that isn't consistent over that period, then you end up spending a lot of time potentially converting data that you don't need to see in the new system. You also need to consider the people aspect because it's incredibly important to think not just about what the controller needs, but what does the company need? Who else needs to access this data? What is the value-add company-wide?
If it's a level of individual comfort because they’ve always run this report and it makes them feel comfortable to have access to that data, then you should consider if it’s important to bring over or not. It's not really worth it for that one individual. We can give them access to a database where they can still query that information, but we don't necessarily need to muddy the new system with dirty data.
