The work that repeats every week is the work worth automating.

Three honest ones. First, we only take on processes where we’re confident in the projection, so we say no a lot; if a process is too messy or too small to measure cleanly, we won’t pitch it. Second, the shortfall refund requires that the system was actually used as set out in the agreed scope, so we’re measuring the automation and not a process that quietly went back to the old way. Third, we’re a young firm still building a public track record, which is exactly why the risk sits with us and not you: the guarantee stands in for the years of case studies we don’t have yet. You can check who we are on our about page and register entry, and on LinkedIn.

Before we build, we agree a baseline together: the labor hours your team spends on the process times their loaded cost, plus any tool licenses we’ll replace and the cost of errors and rework. From that we set a projected Year-1 saving and write it into the scope. After launch we re-measure the same line items, so the number is transparent and mutual the whole way through.

Then we correct the fee. If projected savings were $40,000, the delivery fee would be $20,000. If verified savings later came to $30,000, the corrected fee would be $15,000 and we would refund $5,000. The system has to have been used as set out in the agreed scope, so we are measuring the automation rather than a process that quietly returned to the old way.

Most first automations are working in 5 to 10 days, not months. We deliberately scope the first build narrow, around one painful, costly process, so you see real output quickly. Speed is part of the model: the sooner it runs, the sooner the savings start.