I once agreed a contract with a Business Angel to work on two of his companies. One of them was loss making and the other was evidently booming but it's processes and internal controls were weak.

There was a story about the latter. Eleven months into its financial year their accounting system developed a bug. The program of action the business planned and followed was to replace the system and recruit a number of  temps to re-input the eleven months of data and attempt to get everything up to date. Ten months later when I joined it still wasn't up to date, the prior year hadn't been closed, the business was running without management information (MI) and had no cash (never a good sign).

The system which developed the bug was Sage Line 50. The audit trail had apparently been wiped. This had been the conclusion of the acting Accountant who devised the plan to replace it. 

I have a natural curiosity. An off the shelf package is very unlikely to develop a bug so severe as to merit a complete system change. In addition the support team at Sage will ordinarily fix a problem if one arises.

Day one I switched on Sage to verify my hunch and sure enough when I clicked into the Audit Trail it was blank, but it was only blank because the search filter was on. When I took the filter off all the transactions were there.  When I ran the verify data routine there were no errors. There was nothing wrong with the system.

The company had actually lost over £200K during this blind period although it believed it was making money.

I use this example frequently to emphasise the importance of.

1. Having a written project plan with gateway reviews, a Prince2 like methodology, that prevents plans being ill conceived and carrying on indefinitely without giving any results.
2. Getting a second opinion if the solution seems too drastic. Hiring me for a half day when the problem was identified would have saved a substantial sum of money and over ten months of efforts.
3. Not doing things in a panic. Take time to use common sense, in this example, the un-likeliness of Sage Line 50 being unfixable.