The case for reshoring your processes – or for never offshoring them in the first place…

What would you do if you had access to twice as many staff at half the cost? This has been the opportunity presented through offshoring work over the last decade or so. The evidence tells us you’d do the same things you do today, only cheaper. You might also do one or two more things that you weren’t able to do before. However, there are drawbacks. By the time you have installed the additional management structures, added the necessary checking and quality resources and offset the inflation and currency risk there’s not so much to play with. Moreover, the sugar rush from the initial cost saving is soon baked in, and you’re back to where you started in terms of operational agility. It’s likely you’ll have installed SLAs and be managing performance using metrics and mechanisms to contain the service. However, according to recent surveys by Horses for Sources, you’ll be left wanting in three key areas:

1) access to more innovation

2) what to do with all that big data you are now getting

3) how to increase automation.

Enter the era of robotic automation, and try asking yourself a new question: what would I do with ten times as many resources at a fraction of the cost? I’ll wager you can’t think that far. Yes, you’ll have a bucket list of things you want to get done today, but now you need to go much further to generate the ideas to fulfil the potential! In short, you’ll need to innovate. Fortunately, there should be no lack of opportunity to do so as you begin to harness the vast data sources becoming available through your own systems, mobile, social media and the web. The dynamic of your operation will need to shift from “getting the work done” to “creating value by harnessing your data and resources to best effect, whilst maintaining control of your future costs and improving control of your risks”.

A recent report by Forrester suggests “Enterprises are shifting to smart process apps to engage customers”, those built quickly on agile technologies, designed to hit a specific purpose and bridge the gap between the systems that engage customers and those that traditionally drive the enterprise e.g. CRM, ERP and BPM. As an example, one of Genfour’s clients wished to improve customer uptake through a promotion process. Evidence suggested this was best done during the customer interaction i.e. when they were on the phone or using chat. The answer was to rapidly develop a web services tool for agents to use realtime to query the CRM and return a result to guide the conversation. Then a Blue Prism-based robotic process is able to fulfil the request once the customer had made their choice. The result is a seamless process, with no extra labour required, performed quickly and accurately to drive customer engagement and retention.

If that represents a snapshot of where we are going, what of the reality today? Clearly there are limits to what robots can currently do. Good solutions such as Blue Prism are emerging to handle processes that have structured electronic inputs and follow defined rules. Increasingly we are looking at tools such as Celaton to handle unstructured inputs and apply a level of artificial intelligence to learn about the data it is processing. This then gives scope to take unstructured data e.g. a customer email, to learn the types of requests and/or information received, and route to the appropriate robot all without human intervention. Beyond we enter a realm of true cognitive responses where machines such as IBM’s Watson apply human-like thinking to respond to vast arrays of data. As we grow in confidence with more and more examples of robotic application, we will extend the influence of tools such as these across our operations.

So what do you need to do to prepare for this?  At Genfour, we see four key areas for our clients to focus:

1) Processes: you will need to know your processes to a very precise level of detail. Very often as we help our clients define their processes, we are drawing the rules out of the heads of Subject Matter Experts. We have to deal with the inconsistencies and variances as we go before getting to an “agreed” position, the one the robot will then follow relentlessly. You can’t blame the robot for not doing something right if you haven’t told it the right thing to do!

2) Data: understand your data sources and what plans you have for them. Your systems and data feeds are all in a constant state of flux. You need to anticipate future changes to plan for this. Moreover, there will be choices: for example, do you push a change back to a customer to enter data differently or introduce a tool to handle the data format you are receiving. As robots become more capable, some of the reengineering applied through Lean techniques will be less necessary. You will also be able to handle less structured data feeds such as social media and emails direct into robot workers.

3) Technology: clearly the march of the robots will introduce and employ more technologies and tools. You will need to make sense of these to assemble a preferred toolkit to handle the types of work you have. Then there is the burden of maintaining them into the future, and to access the skillsets and people you need to use them successfully.

4) People: we’ve all discussed the direct impact of robots on the roles they are replacing. An even greater challenge is to assemble the skillsets you will need in your remaining workforce to adapt to and prosper in this new world. The change management challenge here has been totally underestimated thus far. The future will demand innovators, tenacious problem solvers and committed improvers to systemise your operations.

So where to start? In the majority of organisations, a compelling case can be made to get underway by automating your rules-based processes, using a tool such as Blue Prism. To achieve an optimal ROI, Genfour automates 80% to 90% of the caseload for a typical process, which leaves the remaining 10% to 20% exceptions to be dealt with by humans. Overall the automation will yield approximately an 80% reduction in labour. So for a process that required 10 human FTEs, post automation it will require 2FTE. The robots pick up the rest. Onshore, a very conservative human FTE will cost $30-40k fully loaded and a robot FTE equivalent somewhere between $5-8k, depending on productivity. So a 10FTE onshore process costing $400k per annum will realistically cost less than $144k per annum – and potentially much less. That will beat an offshore price point (especially when the cost of management, onshore component of work, staff attrition and training, cost of errors, inflation and currency risk are all taken into account). Importantly, getting underway will commence a change of mindset in your team, developing the areas of focus outlined above.

Now, coming back to the original question asked regarding the case for reshoring your processes, or not to offshore in the first place? I don’t see this debate as being about which shore to locate the work. I believe it will be more important to locate the work at the heart of your organisation – close to your customers, your products and partners – and with control close to your management. Only then will the flow of information derived from your robot workforce, and the requirements of your business be understood well enough to innovate as you will need to.  So that’s where I’d put it, wherever that is!


If you have enjoyed this content then please follow Genfour on LinkedIn and Twitter

This article was produced by James Hall, MD of Genfour, for the ‘Age of the Robots’,  feature by Outsource Magazine issue #35. The full cover story is available here –