I attended Shared Services & Outsourcing Week in Dublin recently and sat in on a session called “Increasing Automation to Optimise Existing Technology”. Part of the session focused on improving the cash collection part of the shared service provision for
a global healthcare company. The firm presented a case study around the receipt of payment from the Spanish government for 45,000 line items worth €40m. Matching cash to invoices for all items presented a major challenge to the SSC operations. Allocation of
the resulting cash through the existing process would have taken four people three months, or ~200 man days at 225 items per man day.
The solution? A major BPO provider developed an Auto Hot Key (AHK) solution to automate the required matching – resulting in a 90% match rate and 10% exceptions to work on manually. This sounds impressive, but a 10% exception rate means 4,500 items to work
on manually – or 20 man days, which still sounds like a lot of manual effort without a proper automated matching platform.
For those that don’t know, “Danger, Will Robinson!” is a catchphrase from the 1960s’ American television series Lost in Space. The robot, acting as a surrogate guardian, says this to young Will Robinson when the boy is unaware of an impending threat. So
why the quote from Lost in Space? The AHK sounds like a user developed application (UDA) that would be better served by an enterprise-wide product. The risk is that AHKs quickly proliferate, leading to over reliance and bringing with them the control issues
surrounding UDAs as a whole.
This is reminiscent of the use of UDAs within investment banks. When these firms encountered growing volumes of equity / cash transactions in their emerging service centres they introduced software products to provide control, replacing existing UDAs. However,
they continued down a path of UDAs for lower volume, non-standard (and often more complex) financial instruments, a path they have largely followed to this day. As a result, non-standard instruments remain error prone, poorly controlled and heavily reliant
on UDAs. The consequence? An inability to then scale the operation significantly because increased volumes mean increased risk that cannot be addressed properly, or at all, in the existing UDA control environment.
The right answer for both investment banks and corporates is to select productised software control from the start. It’s not too late to change, but long term issues can arise when a tempting solution that is rapid and cheap to deploy (such as Excel, Access
or an internal IT build) becomes a key element of your business automation.
In terms of the automation being undertaken by the business as a whole, I thought the below three take-aways were interesting:
- Undertake a thorough process review first: automation may not be the first choice
- Automating elements of what you have is now a viable alternative to total process replacement
- The robots are coming!
I guess you just have to choose the right robot, or risk being lost in space!