Four questions you should be asking about robotics
Been hanging out under a rock? If so, you may have missed the industry buzz about using “robots” to drive out increased efficiency from manual processes. Definitely a catchy concept (humans truly are annoying) and a potential game-changer.
So what do you need to know? I suggest breaking this down into four questions that will help shed light on the opportunity.
- Is robotics new?
- Why now?
- Does it matter?
- What will change?
Let’s take a look at each of these questions in turn.
Is robotics new?
Easy answer: no, not really. However, it is starting to become “improved.” But “new and improved” is more of a consumer marketing ploy than something sophisticated B2B marketers trot out to wow mid-management – so we end up with just “new.”
In reality, the concept is basically inspired by “macros” that matured in the late 1980s to mid-1990s – some of you may remember doing fancy combinations of spreadsheet calculations and print-merge to save dreadful time spent cranking out repetitive tasks in Microsoft Office. Same idea here, but these macros can run across your entire computer and applications suites, creating spreadsheets to do some calculations, updating databases with the results, managing security access across systems, etc: basically a layer of rules-based activity automation that wraps around an existing environment.
To be clear, these robots are not physical, walking machines distributing printed reports around the building from the mail room. Nor are they savvy artificial intelligence-enabled beings. In fact, calling it a “robot” is a bit disingenuous when all we are really referring to is automation through software. (To be fair, with all the “wizards” running around inside my laptop, I can probably accept the concept of some “robots” moving in as well.) As result, robotic automation is about creating increased efficiency and predictability. It does not make the service “smarter” – which is the potential analytics affords us as it matures… and automation combined with analytics will be incredibly powerful.
Although this technology is not fundamentally new, I see two primary reasons why it is taking hold now.
First, the labour arbitrage milking cow may still be producing, but the output won’t be tasty forever, and we need some other source of value to nourish outsourcing relationships. The need and the ability to focus on automating rules-based activities is coming forward to fill the value void.
Second, the technology has matured… small things like non-technical managers being more comfortable with concepts like automation, user interfaces being easier and better, greater commonality across systems and interaction feeds, etc. So the technology is ready and easier to use and our organisations are better equipped to embrace it.
Does it matter?
Heck yes! The amount of manual work that can be automated is significant – I speculate it can eventually be equal or more than the labour and cost impact of offshoring. Some types of work are fairly obvious, other types are not so obvious. These more complicated types require significant process redesign to encapsulate the activities which truly require a human into appropriate modules vs. fragmenting throughout a process. And some exceptions in a process will not be possible to find a way to deconstruct and automate.
As a result, implementing robotic automation requires some real work – plus periodic updates as the process changes. A process that is too unstable may not be a candidate for much automation.
These factors then create scale requirements to justify the investment to implement and maintain the automation tools. However, as organisations implement more and more of these, this should become easier, with more components being transferable and new systems and processes being built from the beginning to enable automation.
Net-net, some painful, detailed work, but a big prize in some situations. Plus no jet lag.
What will change?
On the surface, robotic automation would seem to reduce the case to outsource since an organisation no longer needs to access the same volumes (and types) of labour. I suspect some organisations will indeed alter their sourcing strategies to reduce their reliance on outsourcing. However, I also think outsourcing service providers have a head start on automation insights and skill sets, plus some work will still need humans, which providers may be best positioned to make available. So, expect some confusion and disruption to sourcing model strategies and the service provider landscape (driven by evolving value propositions).
You should also expect to hear some interesting conversations about pricing. I have already heard at least one service provider suggesting it should still be paid per seat for their “robots” (i.e., the equivalent worker that automation replaced… one way to address attrition…). I can’t see that happening very often, but it will further emphasise transaction-based pricing (and less of a cost plus pricing) and in some cases gain-sharing.
The final change to anticipate is further movement towards end-to-end process models. End-to-end is much easier said than done, but when the discussion between different organisations becomes about how they can both capture efficiencies by changing the process, the internal resistance will subside in some cases. Interestingly, this might mean that some end-to-end processes could be implemented with less disruption to traditional organisations – not clear to me that that would be a good thing.
The time for increased automation in the corporate back office seems to finally be upon us – how fast and how far are still question marks, but the direction is clear. Just be careful to not upset the new automation robots: they may someday be smart enough to start taking on judgment-based work! Which is where high-end analytics is seeking to take the next stage of the story…
To read other sections of our article ‘Age of the Robots’, from Outsource #35 (Spring 2014), click here.
About the Author