When evaluating an automation solution, tangible cost benefits are often difficult to quantify. There's the obvious things, like reducing printing check costs, transitioning to ACH instead of checks, and the reduction of hard costs that relate to shipping, but many of the main benefits are more ‘soft costs'.
I've had clients ask me for help with an ROI many times over the years, and I thought it was finally time to put what I have learned into a tangible ROI calculator.
In reading The New Economics by Dr Deming and applying his total quality management ideas into automation has been a cornerstone of everything we do at Fidesic, and applying those principles of quality into an ROI tool seemed like a logical place to start. While this book was written in a different error, and is mostly focused on manufacturing many of the principals apply in an AP system. How do we constantly improve the system of AP?This is perhaps the easiest to quantify: time spent entering data can more or less be completely eliminated. Taking all factors into account here, we've found that actual time spent entering an AP document into any ERP system can take on average 3 minutes per invoice. There are some economies of scale here, but entering an invoice often involves navigation through menus, and taking the time to review approvals (beyond the simple act of entering the data). With Fidesic OCR, this time is completely eliminated.
This becomes a bit more complicated, as time spent approving is a bit more ‘soft' than time spent entering. The time sink that goes into approvals will depend on the level or organization of the individual, and the ability to have all reference materials available while making the approval. In a standard paper based process in my experience, this takes on average 2.5 minutes per invoice. If there is required GL coding, and referencing other materials, this estimate is likely to be quite low.
The other main factor with invoice approval is that it is multiplied by the number of approvals necessary per approval, and we find this to be the most important factor in how much we can save people using Fidesic. If you have multiple stages in approval, cutting each approval into ¼ the time really starts to add up quickly, especially if each user has other management tasks they should be attending to.
This is another ‘soft' cost that can drown an AP team if the process isn't organized properly. AP has to spend time manually routing invoice to approval, and then filing an invoice upon it's return to AP. In a best case scenario, this will also take 2.5 minutes per invoice, and this cost can be completely eliminated though the Fidesic AP automation platform.
Beyond the one-time routing, invoices are often retrieved multiple times at the request of approvers, management, or vendors. We estimate that roughly 10% of invoices before using Fidesic will require some sort of secondary retrieval, and at 10 minutes per retrieval, these costs also add up.
Based on our calculator, how much does it cost to process an invoice? For an average size company with 1,000 invoices and 10 approvers we estimate the cost to process before automation to be roughly $9 per invoice. So that's great, but what does the other ‘industry experts' say on the topic. Estimates vary, but most estimate the cost being between $10-17 per invoice processed.
So yes, our estimate is a bit conservative on that scale, but I feel that makes our tool an even more compelling argument for Fidesic as a platform. We're not inflating costs of invoice processing today to justify charging exorbitant fees on our end, we want the people handling the analysis to be able to see with confidence, that Fidesic is truly going to save them time, and in term money.
So what does our Tool say after Fidesic has been implemented, that same company would be spending roughly $2 to process invoices through our system, and that includes our fees.
I encourage you to go check it out and run the numbers yourself. We're constantly looking to improve this tool to reflect reality more accurately, so if you see something that you think we can improve, please let us know. Happy analysis!