A colleague named Oksana and I were working on a three-year revenue forecast for a small retail client. We used Fathom, which is a solid tool, and let the platform generate projections based on historical data imported from Xero. We reviewed the charts, liked the trajectory, and sent it over.
The meeting that exposed the gap
The client noticed the model assumed a 6% annual growth rate. That was the software default, based on their past two years of revenue. The problem was that one of those years included a government subsidy that was not going to repeat. Fathom had no way of knowing that. We had not adjusted the assumption manually, and we had not flagged it in the report.
What automated tools cannot do on their own
Tools like Fathom, Futrli, or Float are genuinely useful for small business forecasting. But they pull patterns from historical data and apply rules. They do not know your client's context. Any one-time event, contract ending, or market shift has to be accounted for by the analyst, not the software.
Reading the assumptions panel before finalizing any automated forecast is now the first thing I do. Most tools surface these settings somewhere, but it is easy to skip past them when the output looks reasonable at first glance.