I recently had the privilege of attending at the United Nations an invited gathering of individuals active in the Artificial Intelligence (AI) space. The meeting was hosted by UNOPS with a view to consider how AI can be better deployed to support the UN mission. The attendees included leaders in AI and its related fields with the AI leaders for companies such as Facebook, Google, IBM Watson, Intel and IPSoft, along with government entities and distinguished academics and journalists.
At the dinner, a great number of topics were discussed with the United Nations representatives challenging the community to use its collective resources to support the UN’s mission of cleaning up the oceans, feeding the hungry and delivering medicine to the underserved. As the discussion progressed, security emerged as an important topic and that AI has the potential to both make a vital contribution and to be co-opted and weaponized as an agent of attack.
The earlier assumption that artificial intelligence (AI) would impact “routine” jobs first is not holding ground anymore. Indeed, we might be deluding ourselves by thinking that the time in which AI could be the most used interface to engage with technology systems is far in the future. But I’m getting ahead of myself. Let’s look at what’s happening today in the creative sector.
IBM Watson was used last year to create a 20th Century Fox movie trailer. Adobe Sensei is putting the digital experience in the hands of non-professionals. Google is leveraging AI with Autodraw to “help everyone create anything visual, fast.” Think about anyone, any one of us, taking a picture and then telling photo editing software what to do. No need to work with complex brushes, paints, or understanding of color patterns.
AI and creative talent
This is scary for creative people such as graphic designers, digital artists, and others who may consider artificial intelligence a job killing replacement of their skills, despite pundits’ claims that it will “augment” human capabilities. Antagonists proclaim that AI systems will at best reduce the overhead with which creative people deal. But removing overhead is just the first step; the next step is surely the creative. More so, AI systems can ingest so much data, and will increasingly rely on unsupervised learning to correlate so many behavioral traits to create compelling creative content that a human creative artist cannot possibly fathom or understand. Thus, these artists will begin leveraging AI systems to create exceptional, “unthinkable” user experiences that have no “baseline” of reference, but soon may be replaced by these very systems.
AI and businesses
Businesses have always struggled to hire highly skilled creative professionals, and have paid through the nose to secure and retain them. That they will be able to leverage artificial intelligence to take charge of creativity to drive messages and communication to their end users, rather than relying on creative experts, will make them extremely pleased. As the intent of most AI systems is to enable non-specialists to perform many tasks by “hiding what is under the hood,” businesses might not need as many specialist human creative skills.
However, despite this seeming upside of AI systems, their under the hood nature will create problems of accountability. The complexity of their deep learning and neural networks will become such that even the teams developing the systems won’t be able to provide answers to specific decisions they make. Thus, if something goes wrong, where should the blame be placed? With the creation team, or with the AI system itself? The system won’t care – after all, it’s just a technology – and the team members will argue that the system learned by itself, far beyond what was coded, and they cannot be held accountable for its misdeeds. This is scary for businesses.
AI and the impact beyond business
Imagine the impact this will have on the society. Although you can track back how any other technological system arrives at an answer, AI systems that are now supposed to run not only social infrastructure but also much our entire life won’t be accountable to anyone! This is scary for everyone.
I don’t want to add to the alarmist scare going on in the industry, but I cannot be blind to successful use cases I witness daily. People will argue that AI has a long way to go before it becomes a credible alternative to human-based creativity. But, the reality is that the road may not be as long as is generally perceived.
There’s no shortage of books, news articles and comments in social media about how artificial intelligence (A.I.) is shaping our future. Although it’s still blazing a trail, we’re on the brink of A.I. disruption that will change all industries and society at a very deep and fundamental level. I believe it will be one of the next great wealth generators.
My optimism about A.I.’s growing potential arises from many successful use case examples as clear evidence that A.I. is now getting the scale, maturity and the ecosystem in which it can be effective. Although A.I. has been developing for 20 to 30 years, it’s gaining enough elements necessary for a supporting ecosystem.
Major advances in Artificial Intelligence (AI) technology are happening rapidly, and many organizations are excited about the possibilities AI presents. However, some successful companies fail in their innovation efforts to create new value by leveraging disruptive technologies. Others haven’t yet embarked on this innovation journey because they lack use cases. My advice: First define the strategy you’ll use to create value through AI. And Facebook is good role model when it comes to AI strategy.
Your strategy for value creation needs to include how to maximize and expedite the development process. This is the foundational core of Facebook’s strategy. I recently attended a dinner focused on AI at the United Nations. Hosted by UNOPS, attendees included distinguished academics, software companies, government entities and AI leaders in companies successfully using AI.
The business environment in which today’s IT service providers are operating is one of the most challenging in recent times. A host of buy- and supply-side factors are impacting the prices they can feasibly and competitively charge their enterprise clients in the U.S., and their margins are being constricted at every turn.
On the buy-side, ongoing commodity slowdown led to overall softening in the global services market in 2016. Uncertainties created by Brexit in mid-year and the U.S. elections in Q4 delayed decisions on new sourcing contracts and temporary cuts in discretionary spending in SI type engagements.
The quantum of large application outsourcing (AO)/systems integration (SI) deals (>US$100 million annual contract value, or ACV) as a percent of total deals fell from 3.3 percent in 2015 to a low of 1.7 percent in 2016, reducing the pricing cushion typically afforded by large deals. And because enterprises continue to maintain a portfolio of preferred AO vendors to foster price competitiveness and innovation, resulting in a price war for deals, the average ACV in AO deals dropped by ~20% in 2016.
Most enterprises are optimizing their portfolios of contracted relationships to reduce overall TCO by improving nomenclatures, rates, service levels, T&Cs, productivity, etc., leading to a dip in realized revenue per FTE for providers.
Additional downward pressure on realized revenue per FTE has resulted from an increase in brownfield automation, especially in compete situations and second generation renewals. And renewals fell sharply, from 55 percent in 2015 to just 27 percent in 2016, driving price wars among providers.
On the supply-side, although resource utilization increased for Tier 1 service providers from ~80 percent in 2015 to >82 percent in 2016, it is beginning to max out as a delivery optimization lever. Consequently, providers are trying to achieve higher efficiencies and sustain margins via better project planning, DevOps, agile staffing, and proactive use of automation.
There is extreme competition in most rebid and re-compete situations, which has led to an overall decline in pricing. We saw an average dip of 1-2 percent in AO /SI FTE rate cards, but bigger dips in overall account-level TCVs. And per rate cards, some enterprises have pushed for single onshore rate card that doesn’t delineate between local and landed resources, leading to cheaper onshore rates. That said, the new U.S. government may push for more onshore hiring and localized presence, including sanctions on landed resources. This may push onshore rates higher, marginalize the landed resource model, and put additional margin pressures on service providers in the second half of 2017.
All this paints a pretty gloomy picture for IT service providers. However, they have started pivoting towards a digital first future, which can help stem their margin and profit erosion, and reverse the worrisome growth deceleration. Most are growing their top line and/or capability portfolio inorganically. Most are also investing in and pitching automation capabilities in a bullish manner. While this may led to a near-term cannibalization of their traditional offerings, in the medium- to long-term it will help sustain their margins in a price competitive landscape.
Do you believe that a digital first pivot will help service providers get back to double digit growth rates?
In the heat of battle in the services industry’s rotation from labor arbitrage to digital, Genpact made a significant move today that signals to everyone it’s playing to win. Genpact announced it signed an agreement to acquire Rage Frameworks, a leader in enterprise Artificial Intelligence (AI) and automation technologies and services. Genpact moved the cheese.
Three aspects of Genpact’s acquisition of Rage are especially significant.
Serious Commitment: It’s apparent that Genpact recognizes the future of services will be digital. The global services industry is witnessing unprecedented deceleration. At Everest Group, we closely track the top 20 service providers. As illustrated in the chart below, the labor arbitrage-based businesses collectively stopped growing last year and 21 percent of the industry growth is now in businesses with a digital focus.Among the top 20 providers, growth for arbitrage-first providers actually shrank last year. As we look forward two more years, we think it will further decelerate and go to just under two percent.Accenture has rotated into digital faster than other providers, and already has the highest percentage of work in digital. It’s the biggest and is blowing other providers away. Of note, it’s the only provider that has succeeded in growing its margin while moving into digital. As I’ve recently blogged about the dilemmas at Infosys and Cognizant, for example, the rotation into digital stresses providers’ margins. The faster they grow in digital, the lower their margins are, which is very inconvenient.“Change is inevitable, but growth is optional,” wrote John C. Maxwell, leadership expert and best-selling author. With its Rage acquisition, Genpact bypasses the dilemma other providers are facing and demonstrates its seriousness and its commitment to growth in the digital space.
Accelerating clients’ digital transformation: Everest Group’s research group conducted a study of 132 “best reference” clients of top service providers. Our study found 48 percent of clients are unhappy and 25 percent are very unhappy. A top reason for their dissatisfaction is providers’ capability of helping them with a digital restructure.Rage Frameworks presents an exciting set of technologies that are immediately applicable to Genpact’s existing client base. Leveraging Rage’s no-code development AI platform in cognitive computing, enterprises can gain real-time insights for mission-critical functions, simplify automation, manage risks better and gain competitive advantages. For the last 18 months, Genpact and Rage successfully partnered on strategic client digital engagements including a large global insurer, a global consumer packaged goods leader and several large financial institutions. Their combined capabilities will help clients drive digital transformation at scale and accelerate clients’ digital journey.
Rebranding: The Rage acquisition also enables Genpact to rebrand itself as a digital company in the broader marketplace, not just an arbitrage service provider. Moreover, the AI capability is quickly becoming mainstream for leading enterprises as it enables organizations to change the way work is done and enhance their value proposition and competitive advantage.
Requirement for Digital Rotation Success
When an arbitrage company such as Genpact thinks about its rotation into digital, it must focus on managing three constituencies: shareholders, internal constituencies and customers.
The Rage Frameworks acquisition helps Genpact manage across all three constituencies, as follows:
It signals to the shareholders that Genpact is serious about rotating into digital, and it’s joining companies like Accenture, which is the leader of rotation into digital.
It equips Genpact’s internal organization with the tools and intellectual property to drive the provider’s transformation into a digital services leader.
It helps reposition Genpact with its customer base in Artificial Intelligence and cognitive computing – a very important and quickly growing emphasis in digital capability.
Genpact’s bold move is important to watch. How many other arbitrage providers will follow this path of serious investment to accelerate their journey to become digital-first service providers?
Consider this situation: the head of operations for the finance function at a leading insurance company had heard from both his BPO provider and several technology companies claims of robotic process automation’s (RPA) ability to deliver significant operational efficiencies and cost savings. He decided to take the plunge, obtained leadership buy in, made elaborate RPA implementation plans, set the investment rolling, and finally directed his team to implement RPA across the function in all geographies. A year after deployment, he was still struggling to see significant evidence of the promised benefits. Have you heard stories like this before? Perhaps experienced it in your own organization? Unfortunately, it’s all too common. Various factors contribute to the glaring gap between expectations and reality. They can be broadly categorized as follows:
Process – This includes lack of process standardization and an improper process viability study for automation. For example, process-specific differences due to delivery location-specific nuances require greater RPA tool customization. This drives costs higher, and can significantly impact the automation ROI and expected benefits.
Delivery – This includes factors such as service placement across locations that can affect the automation levels in a process. For example, if FTEs are fragmented across different locations, they cannot be released because of minimum FTE requirements to service those locations.
Governance – Similar to any major transformation initiative, success with RPA depends upon various governance specific factors like buy in from IT, organization alignment, etc. For example, we have seen IT departments causing considerable delays in providing the prerequisite security clearances for RPA environment setup and deployment.
Automation assessment, or lack thereof, is the most critical factor
The paramount contributing factor to a successful – or unsuccessful – implementation of an automation project in your enterprise is automation discovery, i.e., the assessment of automation potential in your organization’s unique environment. In all too many cases, BPO or RPA tool providers look first just at process viability, e.g., if the process consists of rules-based, rather than judgment-based, activities, and then base their assessment on experiences with other clients or on a pure numeric automation benchmarking exercise provided by an analyst firm. However, mere replication of the same RPA tool across a similar process will not necessarily deliver similar automation benefits.
Our recommendations for enterprise buyers on how to help optimize their RPA investments and achieve the potential ROI include:
Make sure your RPA vendor or BPO service provider addresses topics specific to your particular environment, including an assessment of impact of a fragmented technology landscape, regional language-specific nuances to be considered, etc.
Obtain a robust automation benefits benchmark mapped to your organization’s context, and safeguard yourself, if possible, with contractual obligations
Set the right expectations with internal stakeholders on potential benefits
Our primary recommendation for service providers looking to satisfy their clients’ expectations is to leverage a comprehensive automation benefit benchmarking model that is based on benefits delivered and the underlying context of RPA delivery. This will help them not only better estimate but also realize the potential benefits on their clients’ behalf. Service providers that correctly estimate, communicate, and deliver these benefits consistently will ultimately have more RPA success stories to tell.
Do you work for an enterprise or service provider that has implemented RPA? If so, our readers would love to hear about your automation journey experience, and if/how you were able to achieve the intended benefits!
In his 1970 book “Future Shock,” author and futurist Alvin Toffler made the argument that the modern world disorients people as it creates so many overwhelming changes that we are unable to handle them. Almost 50 years after the book was published, I was struck by Toffler’s argument during a recent client engagement in which we were helping an enterprise identify virtual agents/chatbots for its customer-facing processes. All of the bots contained a healthy dose of artificial intelligence (AI), and each one was trying to push the envelope.
Is AI starting to overwhelm people to the point that they may get frustrated with developments they cannot fathom or use?
Every day we see and read about new use cases that “wow” us. We are amazed and bedazzled by advances in AI. And some are becoming increasingly commonplace in the consumer arena…just think smart homes.
On the flip side, there have been instances in which consumers have found dealing with these omnipresent home devices scary and frustrating. Humans have already strongly voiced that they don’t need bots to shop. And, feeling the need for peace in their home, they have switched off many of their home assistance devices.
Some may argue that the technology industry, driven by the high intoxication from the ivory towers of Silicon Valley, is getting way ahead of the people who are expected to be the eventual consumers of these technologies. The amount of new AI research and products coming every day out of these factories is mind numbing. A significant number of such products may not have any immediate utility, but they do indeed demonstrate the far-reaching power of such advanced systems.
There is an unending scare around AI, cognitive, and other advanced systems taking away jobs from human beings. In the case of virtual reality, people are entranced by engaging with virtual objects as if they are real. It’s fun, until they realize the negative impacts it can have on their day-to-day lives. And, instead of assuaging such fears, the technology industry continues to create use cases to replace human tasks with robots.
From an enterprise perspective, organizations need to proactively create an AI adoption strategy for their business. Though most now have some vision around using AI technologies, frighteningly few are preparing for the massive change management aspect. Their employees must be comforted around the impact AI can and will have on their lives. Indeed, the significant disruption AI technologies can create within a business context may require a very different approach than other technology adoption we have ever witnessed. Technology vendors need to focus on how AI-enabled systems are assisting or helping human beings. The use cases need to be very precise, clear, and friendly, not overwhelming and complex, which they currently are.
The problem is not AI technology. The problem is the way it is being introduced, and the hyperbole around it that may end up overwhelming a significant portion of the human race, leading to eventual burnout. We are humans, and should create technology for humans. If the very technology we create results in alienating a large percentage of us, we will have failed as a human race. AI systems need to be leveraged for enhancing human lives, not for creating technology marvels that overwhelm people and create the future shock.
Recently, as guest speaker at an event for senior lawyers, I looked at “Artificial Intelligence (AI) and the Shape of Things to come”. I started by asking who was using artificial intelligence. A few hands went up but of course, it was a trick question! I held up my mobile phone and pointed out that everyone using Google or mobile assistants, such as, Siri and Galaxy, is using some form of AI – and it neatly demonstrates that AI has arrived in all our lives in ways we have not even realised.
AI is working its way into all aspect of work, business and leisure. The likes of Amazon Echo and Alexa have brought AI to the home while businesses have started to use AI to handle some of their core functions. Examples include processing invoice payments, insurance claims and customer complaints. In the professions such as legal services, some firms have deployed AI that decides what paragraphs to include in legal contracts. On another front, AI is being used in law enforcement, helping police forces uncover fraud.
AI is here and is touching our lives one way or another, sometimes without us even knowing it. So what is it and what kind of benefits or challenges does it bring? I am not an AI scientist and in this blog, attempt to shine only some light on this vast and fast developing topic.
A sea change is starting because of digital technologies. The impact as companies apply these technologies to their business will be massive – much bigger than the Industrial Revolution with the invention of the loom for manufacturing clothing and Ford inventing the production for manufacturing automobiles. Everyone has been talking for some time about how big an impact these technologies will have on the services industry. But there is a new factor now that makes the potential impact even more significant: the protectionist activities driving companies to step back or pause in globalization and offshoring. I think the services industry would be foolish to ignore the potential of this greater impact. Let’s look at where businesses are headed.
There can be no denying that the stakes have been raised and barriers are being put in place to make globalization and offshoring less acceptable and expensive. In Europe, it is evident with the Brexit bill and the UK opting to leave the EU. In the US, protectionist barriers are starting to be executed through proposed changes to immigration law and H-1B visas, tax reform and potential border tax implications, and reputational risks arising to companies from government entities or disgruntled employees and vocal press entities. The result: companies are paying more attention to how to do work onshore without suffering negative cost impacts.
By investing in digital technologies such as Robotic Process Automation (RPA), cognitive computing, automation and cloud, companies can drive cost improvement by dramatically improving the productivity of their workforce. In many cases, they can achieve cost improvement even greater through improved productivity than through labor arbitrage and thus offset impact of not sending their work offshore.
Of course, service providers also can use these technologies to improve their own workforce productivity to offset the potential of rising costs from immigration and H-1B visa reform in the US.
Our market data shows leading providers in the services industry have been looking at digital technologies and associated digital models well before this step back in globalization. Our tracking of service providers clearly shows the traditional services (labor arbitrage, offshore factory model, remote infrastructure management and asset-intensive infrastructure) grew by only .1 percent last year. Almost all the growth in the IT and business process services market came from new digital offerings – which are currently growing at over 18 percent a year.
Although the trend in digital services has already been growing, we believe the current climate discouraging globalization and offshoring will further accelerate the adoption of digital models. This will force the current shared-services structure. It also will force the provider community to fundamentally change their business models and the way they currently structure their business to deliver services.