Outsourcing Contract Practices in the Age of Intelligent Automation

HfS Research and the Sourcing Interest Group (SIG) recently established a partnership to deliver global research and insight to help our respective members and subscribers successfully navigate change. At SIG’s Fall Summit, Elena Christopher co-presented on the topic of the impact of Robotic Process Automation (RPA) and Artificial Intelligence (AI) on outsourcing agreements with Brad Peterson. Brad is a renowned outsourcing attorney with the law firm of Mayer Brown LLP. Together, they melded the market perspective on the potential of intelligent automation for transforming business operations with the practical concerns of successfully contracting for it. This POV summarizes their perspectives and provides actionable recommendations for incorporating the emerging duality of labor and automation in outsourcing agreements. A special thanks to Brad for co-authoring this POV.

Intelligent Automation, including RPA and AI, are poised to transform outsourcing. This transformation will dramatically reduce cost and increase capability, and will help buyers achieve the HfS Triple-A Trifecta of Automation, Analytics, and Artificial Intelligence. There will be winners and disrupted losers among the outsourcing provider community. Outsourcing buyers should act now to capture their share of the benefits and address the risks.

Intelligent Automation Is Changing How Outsourcing Is Done

After considerable time and hype, RPA and AI are real. Every enterprise needs a strategy. The long-term impact as we embrace digital transformation is a reimaging of business operations and the processes and functions that support them. The HfS vision is the Digital OneOffice – a real-time customer centric enterprise fueled by robotic process automation (RPA), artificial intelligence (AI), and smart analytics leveraging the output of RPA and AI. These functions work together with people and other infrastructure and applications to break down silos and build integrated digital business operations. As the silos break down, so too do the organizational constructs such as front, middle, and back office. Labor arbitrage-based outsourcing models will be forever changed by intelligent automation.

While we are on the road to the Digital OneOffice, here are some facts about RPA and AI to help define the technologies and their role in the journey:

• RPA – this is the current “it” automation technology as evidenced by uptake and activity focused on improving process efficiency. RPA software is a programmable tool that allows businesses to tailor complex automations to reduce the human labor required in many repetitive, complex, ruled-based actions and processes. Many enterprises and their service partners are engaged in RPA initiatives, with decent depth emerging around F&A and contact center projects. However, HfS research indicates that much of the current use of RPA is focused on cost savings by removing labor from processes with little to no process rethinking or new process creation. We expect that its potential is considerably greater when used with end-to-end process redesign.
• AI – this refers to the simulation of human thought processes across enterprise operations, where the system makes autonomous decisions, using high-level policies, constantly monitoring and optimizing its performance and automatically adapting itself to changing conditions and evolving business rules and dynamics. It involves self-learning systems that use data mining, pattern recognition, and natural-language processing to mimic the way the human brain works without continuous manual intervention. Unlike RPA, AI systems are trained, not programmed. HfS believes AI holds broader potential than RPA, but is in much more of a nascent stage with technologies and use cases still being identified. AI is not a single technology but a continuum of tools, which makes it difficult to define and articulate its uses and benefits. AI holds great potential for solving business problems often through changing or reimagining business processes. This is part of why it will have strong long-term benefits and ultimately have a much larger effect on the outsourcing market than RPA.

In terms of uptake, HfS’ recent State of Automation study indicated that four fifths of respondents are in either stage one or stage two of our Intelligent Automation Maturity Model. Initiatives in these stages are generally focused on cost-savings (taking people out of processes) or process effectiveness (improving an element of a process). From an outsourcing perspective, initiatives in these stages are often shedding resources in manual roles, such as data entry, but they are also adding new roles such as bot programmer or exceptions management to help oversee the automated elements. The use of intelligent automation requires a change in how outsourcing engagements are crafted and managed.

Exhibit 1: The Intelligent Automation Maturity Model

Source: HfS Research “State of Automation 2017”; Sample: Enterprise Buyers = 400

Service Providers Need to Embrace Intelligent Automation or Risk Disruption

Growth rates for offshore outsourcing have been dropping steadily for over a decade. Labor arbitrage is a fading business model. RPA and AI hold the potential to help outsourcing providers become true partners with their clients but also have the potential to disrupt outsourcing providers who fail to act.

There is a substantial amount of hype that intelligent automation will substantially reduce the need for people in IT and business process operations. HfS’ recent study on the impact of intelligent automation on services jobs indicates that RPA and AI will drain some offshore locales of low-level, manual jobs. India and The Philippines will feel the biggest impact for these types of jobs. But the story is not only the job loss. New middle- and high-end jobs will be created focused on programming bots, managing bots, addressing exceptions identified by bots, digital process development, Design Thinking, decisioning, auditing, etc. It is entirely possible that the per-transaction savings will be so great as to drive enough work to outsourcing providers to result in net job growth

Many service providers are embracing intelligent automation to improve the value that they offer to their clients and to move away from the commodity labor arbitrage business model. Leading outsourcing providers are touting investments in proprietary RPA and AI technology and partnering with RPA and AI software providers.

Eight Recommendations for Including Intelligent Automation in Outsourcing Agreements

We’ve developed eight recommendations to help enterprises consider and include elements of Intelligent Automation in their outsourcing agreements. The recommendations are outlined in Exhibit 2 and then elaborated on in the subsequent text.

Exhibit 2: Eight Recommendations for Including Intelligent Automation in Outsourcing Agreements

Source: HfS Research & Mayer Brown October 2017

1 – For existing contracts, be proactive about opening a dialog with your current outsourcing providers about their current or potential use of intelligent automation. If they are using any forms of intelligent automation, discuss the impact on their cost and profits. Are there opportunities to mutually benefit on any realized efficiencies or obligations to share those benefits? Have they introduced new risks that your contract was not designed to handle? If they are not currently leveraging intelligent automation, you should request an update on their strategy for embracing it and determine the path to mutual benefit. Delivering the benefits may require renegotiation, but this is a tremendous time for customers to drive the agenda because outsourcing providers are eager to find a way to adapt to these new technologies.

2 – For new contracts, include intelligent automation elements such as RPA and AI as criterion in your evaluation and selection of your provider(s). Providers should actively be promoting innovation beyond antiquated labor arbitrage models. If possible, design the outsourced process and select a provider based on the effective use of RPA and AI.

3 – When analyzing expected benefits, beware the hype and have realistic expectations. While labor elimination is a real impact of RPA, projects are often focused at the sub-process layer – such as automating the overdue invoice sub-process of the broader order to cash function. So, for example, that 40% reduction in labor requirements might be 40% of 20% of the overall scope, giving you an 8% benefit. Thus, the impact may be smaller than expected. Be very specific about stated benefits and challenge reductions that seem inflated.

4 – Change the contract to fit a mix of human and digital labor. Traditional outsourcing agreements are designed with the assumption that the work will be done by people and rely on clauses such as “good and workmanlike conduct” and “adequate numbers of appropriately trained and experience personnel.” For outsourcing agreements involving RPA and AI, consider including clauses such as:

– Visibility and approval rights for use of RPA and AI in outsourcing agreements.

– Obligations for RPA and AI implementations to meet defined specifications.

– Access, license, and support clauses similar to a SaaS agreement where the customer will be an end user of the RPA or AI software or will need to use that software as part of any disengagement.

– Intellectual property rights in work generated by RPA or AI.

– Validation to baseline the service levels before automation and verify that none are degraded.

– Process validation to document what will be automated.

– Clear governance model including having people represent the AI or RPA “bot” in governance meetings.

– Inclusion of change management to facilitate even the simplest of process change so that your outsourcing arrangement does not impede you in changing your internal systems.

5 – Reimagine the service commitments for humans versus automation:

– Change the scope of services to explicitly state which tasks will be performed by humans and by bots, perhaps restructuring the scope to reflect your strategy for internal and external use of bots.

– For tasks done by bots, replace service levels designed for human labor (such as average hold time) with service levels designed for the digital workforce (such as exception processing time).

6 – Address emerging data risks. For example:

– Hacked RPA or AI software could be far more dangerous than a rogue employee because it can work much faster than a rogue employee – and with more data. Add cybersecurity and data privacy clauses directed at the data security risk of RPA and AI software.

– Contracts with digital service providers can, unless carefully reviewed, allow the RPA or AI provider to use customer data (including derived data and insights generated by AI and smart analytics). If data is a valuable asset for your company, invest in adequate contract review to protect it.

– Putting data into AI systems that are shared across companies means that you give the value of that data to the provider (and the other customers) but using a dedicated instance of the AI will mean working with a less-insightful AI. This is a critical choice before starting to use a shared AI because there likely will be no way to remove your data from what a shared AI has learned.

7 – Reprice to reform incentives and share gains. Emerging pricing recommendations include:

– Cost reduction commitments from providers to share the benefits of RPA and AI in labor-based deals.

– Project-based pricing for consulting and implementation of (design and build) elements of RPA and AI implementations.

– Gainsharing or other mutual benefit sharing based on a firm baseline.

– Restructuring pricing to move from FTEs and other input-based measures to output-based measures (such as transactions processed) or outcome-based measures (such as business results achieved).

8 – Consider your related contracts. Some cloud-based and subscription-based services prohibit use through bots. Others may impose surprisingly large charges, including by treating use by a bot as if it were use by every human who works through the bot.

Bottom Line: Update Your Outsourcing Contract Practices to Capture the Potential of Intelligent Automation While Mitigating Risk

RPA and AI are a reality and every enterprise needs a strategy. However, the technologies and use cases are still emerging, with the majority focused on cost reduction as the simplest measure of efficiency gain. As we evolve, buyers need to increasingly focus on defining their digital path through reinventing processes and breaking down legacy silos. Outsourcing providers need to be aggressive and embrace RPA and AI or they will simply fade away. Operations is increasingly about doing more with less and outsourcing relationships need to reflect this. This is a superb time for buyers and their providers to negotiate commitments, options, and incentives in existing and new outsourcing contracts that capture the new value and mitigate the new risks of RPA and AI in outsourcing.