Sumitomo Announces Largest-ever RPA Initiative, but is it really all RPA?

What’s happened?

Major Japanese financial services conglomerate, Sumitomo, has announced an unprecedented RPA initiative with savings claimed to be in the hundreds of millions of dollars. However, while this is a tremendous public endorsement for the potential benefits of RPA, this appears to be a massive corporate restructuring that is using RPA as a catalyst for labor reduction – and we already have discussed (see link) how Japanese firms love to deploy automation to increase productivity and competitiveness. The Japanese firms do not suffer from the same negative connotations.

Why this all sounds a bit over-ambitious

Our research into Robotic Process Automation (RPA) consistently reveals that scaling operations is one of the biggest challenges out there facing stakeholders today. Moving to enterprise wide deployments is orders of magnitude more complex than initial proof of concept or pilots. Getting the right principles and governance in place from the get-go is critical to getting it right. This means the methodologies, the operating model, specific milestones, transformational change and an RPA platform that is designed for productivity, adaptability, high security and scalability.

Getting it wrong has a huge impact on the capability for RPA initiatives to deliver against agreed outcome objectives, and plays a major part in disappointing RPA Customer Experience (CX). It’s why we’re going deep into customers’ RPA programmes now to find out what’s going wrong and what can be done to deliver those elusive business outcome goals, namely massive cost reduction, improved productivity and rapid time to value.

The Sumitomo RPA initiative

Because of the inherent risks to RPA programmes, coupled with corporate paranoia surrounding negative publicity of job losses tied to automation, it’s not surprising that customers are wary of disclosing too much too early. So it’s really breaking the mould to hear the specifics provided by Japanese financial services giants Sumitomo Mitsui Financial Group (SMFG) and Sumitomo Mitsui Banking Corp (SMBC), regarding its ambitious plans for RPA, where the companies are aiming for massive cost savings of $450m over the next three years, and nearly $1bn in the medium-term. 

This is unprecedented ambition in recent times from a services contract – numbers of these magnititude are normally associated with massive corporate restructuring, job redundancies and infrastructure sell-offs.

Blue Prism and UiPath are the RPA tools being used by Sumitomo’s group of SIs and consultancies – Accenture Japan, EY, Deloitte, IBM Japan and PwC. So RPA is clearly just one part of a much larger business change programme ongoing across the organisations.

One of the largest proposed RPA deployments

It’s being called one of the largest RPA implementations worldwide, although ironically we don’t yet know how many robots are going to be used.

We understand it’s a significant contract, and most of the implementation and transformation work is going to IBM. UIPath is doing the front office automation – mainly the desktop RDA – and Blue Prism has also been involved doing some unattended automation in the back office, but SMBC declined to mention them in their press release (see link) which clearly points to UIPath as the prime automation software partner in the engagement. 

According to SMFG/SMBC, the attended digital workforce supports the group’s front-office centric activities, enabling the staff to develop the automation themselves and to work alongside the robot by exerting direct command over it. Complementary, the unattended digital workforce targets all the high volume processes that do not require human touch, working 24 hours per day and 365 days per year to sustain high-throughput, high-intensity processing. SMFG/SMBC has endorsed UiPath as ‘highly usable and scalable’ supporting the initiative during this week’s UIPath conference in New York.

RPA scaling going mainstream

SMFG/SMBC have provided a detailed run down of what’s been achieved to date. The group HQ has been going through an operational efficiency programme to simplify processes and eliminate duplicate processes. Of those remaining some have been earmarked for RPA. Since the start of the programme in April 2017, some 200 roles have already been removed (which equates to 400k hours of value released). By the end of the year, this is expected have risen to 1,000 hours. Within three years the expectation is for this to rise to 3m hours and a staggering 1,500 FTEs reduced.

The Japanese financial services sector is really going for ambitious use of automation. Earlier this year rival Fukoku Mutual Life Insurance announced another ambitious first, using IBM’s Watson Explorer to automate mid-office insurance agent processes, but that was supposedly only taking out 30 heads. AI is lagging RPA when it comes to the realizable potential today, and clearly in terms of scale. But we have no doubt that will change as adoption of broader Intelligent Automation initiatives start to form at the intersection of analytics, automation and AI.

Lots of dissatisfied RPA customers out there

In our view SMFG/SMBC are going to come across some big challenges to achieve these ambitious targets.

We are finding big discrepancies between customers who are getting good results from their RPA deployments, versus those for whom it is failing. Our research shows that just 58% of RPA customers are satisfied with their implementations to date.

Our new research into RPA Customer Experiences, which will be available in the coming months, is revealing some huge findings – misaligned expectations between customers and suppliers – made harder still because of the diversity of customers out there, and huge gaps in understanding the broader change and process re-engineering implications.

When RPA and RDA are being used as tactical tools to take out cost, rapidly, this is where the big disasters are happening.

However, where it is being embedded as a strategic component within a broader Intelligent Automation strategy, with buy-in from the top, and engagement across the organisation, the success stories are much greater. Breaking down political barriers between IT and operations is also a critical part of the picture, because time and again, the need for governance and controls becomes the over-riding factor when it comes to scaling enterprise wide.

CofE’s have plenty of drawbacks

It’s no surprise then that SMFG/SMBC are deploying an RPA Centres of Excellence (CofE) in this case. CofE’s are great ways to centralise the RPA capabilities while wrapping important layers of governance and controls to which organisations have to adhere.

But on the other hand, we’re seeing plenty of examples where CofE’s are equally thwarting innovation and stifling the operation’s teams from getting access to RPA or other IA tools that are going to help them achieve their objectives. CofE’s are typically owned and operated by the central IT function, and report into the CIO. This is at odds with operations managers who report into the COO or CFO. Billing back to the CofE for work, rather than having ownership of the costs, loses visibility on some of the most important metrics relating to ROI.

Likewise, if training and expertise is being supplied by the external SI/consultants, these staff are likely to move on to the next job at some point down the line, taking all their knowledge with them. 

Bottom Line: Massive kudos for RPA going at scale, but this appears more like a massive corporate restructuring using RPA as a catalyst for change

There’s no doubt that right now this endorsement from SMFG/SMBC is great kudos for the RPA vendors aiming to scale enterprise wide – the most jaw-dropping initiative yet that takes RPA to the 9-figure level in terms of perceived monetary value.

However, RPA initiatives, in general, have not nearly met cost savings and productivity targets anything near these touted outcomes.  We’re just not there yet as an industry – sure, many of the more mature deals today are yielding value benefits in the $1m-10m range, once they are being managed effectively, but to jump from these levels to the hundreds of millions is massively far-fetched. 

At HfS, while we believe there are genuine intentions from Sumitomo to leverage RPA to free-up and eliminate human labor, there has to be a much broader corporate restructuring plan, way beyond the digital labor initiative, that will get Sumitomo to these lofty targets. It’s also important to point out that these ambitious enterprise wide deals are already going on in other organisations – for instance we understand Blue Prism is involved in a number of projects of at least this size across the globe. But these companies are much less willing to share the details of their programmes with the wider community – clients, employees and shareholders are all going to be impacted in some way by the expectations being set, and whether they are realized or not.

Keeping quiet is often the easier way to avoid the kickbacks when things inevitably go wrong. Whether naïve or not, SMFG/SMBC’s level of disclosure means it going to be an important test case to assess the success of scaling RPA across the enterprise.