Go Digital or Go Home
Fortune 500 companies across all industries are laser-focused on the pace of innovation, the deployment of new and disruptive technologies, and the urgency to digitally transform their business operations and enhance the customer experience. A successful digital environment is characterized as one that is collaborative, customer-centric, visionary, enabled by technology, driven by data, focused on growth, and most importantly, agile.
The legacy outsourcing model was predicated upon pushing the limits of labor arbitrage (utilizing an offshore, less expensive, highly skilled, educated, and language enabled workforce) to the maximum extent possible. Low price and consistent service level achievement were primary buyer values while innovation was of secondary importance. Given that dynamic, one might predict that the digital journey would detrimentally impact the outsourcing industry. However, the opposite has occurred as companies have embraced deploying Intelligent Process Automation (“IPA”) into their outsourcing delivery environment as a mechanism by which they can increase their pace of innovation; support their strategic objectives; reap the benefits of a digital-friendly environment; and still achieve the cost-savings from the pre-digital era.
What is Intelligent Process Automation?
Intelligent Process Automation combines Robotic Process Automation (“RPA”) with Artificial Intelligence (“AI”) to automate complex end-to-end business workflows that are both rules-based and require intelligent decision-making. While RPA follows a script to automate simple, repetitive, and rules-based tasks by mimicking human behavior, AI simulates human intelligence to understand, learn, and make complex decisions from unstructured data sets. Think of RPA as the digital assistant that sticks to the script and performs simple tasks like data entry while AI is the brain of the operation that adapts, learns, reasons, and solves problems. When the two are combined, the result is significantly greater efficiency, accuracy, and cost savings, which yields the ability to shift the focus of the human worker (“FTE”) to higher-value strategic activities that may require professional judgment to perform.
For the RPA side of the equation, software robots are programmed (the RPA software allows the user to map the end-to-end business process flow in the application) to execute high-volume, manual, repetitive, and rules-based processes and tasks. The robots never get tired or hungry and can run 24×7 as necessary. They can scale with minimal impact and predictability, and yield more timely, quality, and cost-effective delivery than
their human co-workers. Each action the robot takes, which may include accessing external websites to obtain data or conducting quality checks, is monitored and recorded so audit and compliance requirements are more efficient and there is a reduced risk of theft or misuse of information. Finally, robots are unlikely to retire or accept work with a competitor so institutional knowledge does not leave the building via attrition.
The AI part of the equation simulates human intelligence to understand, learn, reason, and make complex decisions from unstructured data sets including texts or images. While AI shares many of the advantages of its RPA partner in the areas of cost, speed, accuracy, and scalability, it is dynamic in its execution as it understands context, can manage variability, and learns, improves, and adapts over time.
What Exactly Has Changed? Can We Just Add a Short Appendix to our Standard Template?
At its core, a legacy outsourcing agreement consists of two key components, a monthly managed services fee which is based upon the number of FTEs that will be deployed to deliver the base services and a service level methodology which contains the quality and performance targets expected to be achieved and against which the service provider will be measured during the delivery term.
Over time, the monthly managed service fee will gradually decrease as the service provider achieves greater productivity and the service level targets will gradually increase based upon the continuous improvement provisions contained in the service level methodology. The size of the base services team (which historically has been composed of 100% human workers located in delivery centers around the world) is dependent upon the baseline transaction volume (expressed in terms of incidents, tickets, or calls) it will be responsible for handling. To the extent the actual transaction volume increases or decreases beyond a contractually stipulated dead-band (generally plus or minus 10 percent of the baseline transaction volume), the price will automatically adjust based upon a variable fee structure (generally expressed in the form of Additional Resource Charges/Reduced Resource Credits) stipulated in the agreement. To the extent the actual transaction volume falls beyond the boundary of the deadband on a recurring basis, the parties may need to renegotiate terms.
As referenced above, the goal of the legacy outsourcing model was to push the limits of labor arbitrage wherever possible. The widespread use of IPA will yield a hybrid workforce where humans, robots, and AI work jointly to meet service levels, drive innovation, and deliver value. While robots, AI, and humans can certainly co-exist peacefully, the underlying terms and conditions contained in a legacy outsourcing agreement do not contemplate the deployment of robots or AI into the delivery model. To that end, the terms and conditions in those legacy agreements must be amended to accommodate a hybrid workforce and the changes required may be significant. To facilitate properly capturing the IPA structure, the following questions should be asked:
1. Should the fee structure be changed from a monthly managed service price to a transaction, outcome, or value-based model?
2. To the extent a monthly managed service fee remains, how do you assign value to the AI and RPA components of the base services team?
3. If many of the FTE limitations are eliminated with the deployment of IPA, is it legitimate to expect service levels to be set near 100 percent over the term?
4. Given that RPA and AI work 24×7, is a deadband or ARC/RRC structure necessary to address variances (seasonal or resulting from an acquisition or divestiture or other material business event) in volume?
5. Will decision rights regarding the composition of the delivery team (FTE versus IPA) reside solely with the service provider? Will there be limitations placed on each resource type over the term?
6. Will exception cases that require human intervention be subject to different service levels?
7. How will the deployment of IPA impact the allocation of blame and the consequences of failure (liability limits, indemnity, acceptance criteria, warranty, service level deficiencies, termination, and step-in rights)?
8. What happens if the AI makes a decision that is suboptimal, that yields a breach of the agreement, or that detrimentally impacts a customer relationship?
9. What if the IPA destroys data or fails to comply with data handling protocols which results in breaches of confidential and personal information, security violations, and facility and audit deficiencies?
10. Which party owns the AI, the underlying applications, algorithms, and process automations? Should the digital brain that understands and adapts be treated differently from the digital assistant that follows a script?
11. Will the focus be on best-in-class/industry specific process automations that are owned by the Service Provider so they can be aggregated, anonymized, and fine-tuned to leverage best practices and maximize efficiency or a strong desire for client control and IP ownership?
12. What about third party intellectual property that may be embedded in the IPA?
13. Given that the IPA can be shut down quickly, do transition, termination rights, termination fees, and termination assistance provisions change because of its deployment?
14. Given the pace at which the IPA can learn, is a lengthy transition period necessary?
15. Does the IPA continue to function during a force majeure event? Are there any instances or relief events in which it is excused from performance? Is there an accelerated recovery time objective for the IPA component of the workforce?
16. What happens to the IPA and all underlying tools after an expiration or termination event?
17. Which party is responsible for confirming the IPA will perform in compliance with laws and
regulatory requirements?
18. Which party is responsible for the build specifications, maintenance, and management of the IPA?
19. Does the scope of the benchmarking provision extend beyond price such that a benchmarking condition could be triggered resulting from the failure of the service provider to expand its deployment of IPA and eliminate more human workers?
20. Who will absorb the on-going cost associated with expanding the IPA deployment and eliminating (wind-down and demobilization costs) service provider personnel?
Do We Need to Change our Go-to-Market Strategy?
In addition to amending the legacy outsourcing agreement to accommodate a hybrid workforce, the deployment of IPA will require both service providers and clients to change the approach by which they procure, receive, and deliver outsourcing services.
Service providers will need to:
Target investments, acquisition plans, and organizational changes to reflect that the limits of labor arbitrage have been reached in many geographies.
Deploy new technologies as they become available and focus on innovation—better, faster, and cheaper—over the delivery term.
Adjust the size, skillset, and geographic location of the delivery workforce.
Deliver a more robust set of consulting services to mitigate the possibility of lower margin IPA-led engagements.
Accept higher service levels and greater productivity gains over the delivery term.
Either: (1) Change the FTE-based pricing model to one which is a transaction or outcome-based (per ticket, per server, per invoice, or per claim processed) to maintain current operating margins; or (2) Break the link between the number of FTEs deployed and price by assigning value to the IPA that is consistent with the impact it yields over the term.
Re-train its human personnel to focus on higher value transactions and more frequent customer engagement.
Clients will need to:
Be open to a transaction or outcome-based model. If a monthly managed service fee remains, seek to understand the value being assigned to the FTE and IPA components of the solution.
Negotiate short term agreements with broad termination rights and the ability to lift and shift scope with limited financial impact.
Focus on innovation and the rapid deployment of technology as a key buyer value when selecting an outsourcing services provider.
Secure greater productivity and service level targets.
Demand a preference for automation—why pay for a human when IPA can accomplish the same task in a more timely and quality manner?
Focus less on IP ownership for more generic best-in-class automation so long as a perpetual license is granted that extends beyond any termination or expiration event.
To succeed in the digital outsourcing age, service providers will have to innovate and transform to expand their delivery footprint. Even with those changes, the core principle and foundation upon which a legacy outsourcing agreement is based—delivering the service in a timely, quality, and cost-effective manner—remains unchanged. To that end, a focus on basic blocking and tackling is still mandatory, including a narrowly tailored statement of work that is free from ambiguity and clear on the roles and responsibilities of the parties; a robust list of assumptions or dependencies upon which the service provider’s delivery is conditioned; change control and governance processes that are strictly followed; and acceptance criteria that are objective, measurable, and verifiable.
Innovation has historically been a key buyer value but the service provider commitment in this area has always been ambiguous and difficult to measure. Moreover, the contract terms (specifically the penalties for pervasive performance deficiencies) have dictated a focus on service level achievement first, and innovation thereafter. The advancements in IPA and its ease of deployment have shifted this dynamic. Going forward, innovation will be front and center with service level achievement and productivity being a by-product thereof.
Make no mistake, the legacy approach to placing fifteen percent of fees at risk tied to the achievement of critical service levels and committing to conservative productivity targets will no longer be sufficient. As we enter the next evolution of outsourcing, a substantive commitment to innovation and the high standards of timely, quality, and cost-effective delivery it yields will be required.
