One thing I have learned well is that some things are easy to change later; some are really hard. A real design master might cost you 10% more upfront, but save you a fortune in the long term. Badly designed software may need to be scrapped in 5 years; software with a great design may be useful for 20 or 30 years while you amend it to your changing needs. The financial impact can be major.
A crystal ball would help. Few of us have one. Instead it's a case of understanding the key assumptions built in to the software and either eliminating them or being certain they are assumptions which will stand the test of time.
Platform or language is a common assumption for example. Some Cobol programs from the 70s lasted well in to the 21st century because they were modular and key modules could be retained and wrapped in something else. The assumption of Cobol Forever had been resisted in the original design and the possibility of change allowed for.
All too often senior management doesn't understand software development. Their focus is on meeting budget now. It's the ultimate in short-termism. It's likely possible but the long-term costs to the business could be huge. A skilled designer should be able to swap between technical meetings and the board room where she can explain the trade off between modest savings today and massive additional cost five years down the track.
The rest of the board needs to find such an expert ... And then listen to her.