SOCAP International

The AI Revolution

Artificial Intelligence is set to disrupt and transform businesses across industries, and there will be clear winners and losers.

Automation is helping companies across industries do more with less and offer more attractive services. And many businesses are deriving even greater insights and improving customer engagement by turning their attention to artificial intelligence. While interest in AI is high, its actual application remains low. According to Forrester Research 58% of business and technology professionals are researching AI, but only 12% are using AI systems. That will change, however, and quickly. Forrester predicts that investments in AI will increase by over 300% in 2017 compared with 2016.

The Revolution Has Begun

Opportunities_With_AIThe evolution of AI marks a major revolution. Following the likes of steam, mechanical and information technology, AI is set to disrupt and significantly transform not just the technology space but all industries. And in doing so, there will be clear winners and losers.

Often, it’s the larger, more established companies with bigger budgets and wider resources that are able to jump on new trends. However, AI will operate on a different playing field. Being an incumbent is not going to guarantee success in this revolution. It’s the smaller, nimble companies that could get the edge in bringing innovations to market.

The winners will be those that can adapt and find new ways to apply AI within their specific industry. While most companies stand to gain in some way from the AI revolution, sectors that strive to gain more value from high volumes of customer data have led the charge. Retail, financial services, healthcare and energy are examples of those primed for the next wave of innovation.

They have all orchestrated many successes that illustrate the strengths of AI. However, they also point to potential new uses and where companies could stay one step ahead in future.

Al in Action

AI technology can take many forms, including machine learning and predictive data analytics. Broadly, the current application of AI falls into three main categories.


1. Eradicate banal tasks. 
This is perhaps the largest category and where many existing successes fall. Here, companies are using AI-enabled robotic process automation (RPA) to help eradicate simple tasks such as filling out forms, performing help desk duties or streamlining back-office processing.AI These all require some level of intelligence but take a long time to perform manually, particularly when dealing with high volumes of data.

One example that spans numerous sectors, including retail, wealth management and utilities, is embedding greater automation within customer care and, in particular, call centers. Companies are replacing human agents with virtual agents that can respond intelligently to queries and continually learn and increase their level of knowledge and ability to handle requests.

Another common example is using AI to automate invoice processing, which also improves the payment experience for customers. The technology first scans the invoices to understand the data, then validates them for accuracy. This includes ensuring the charges add up to the total sum, the counterparty names and addresses match internal records and the total is within predefined thresholds. If there are any exceptions, the invoice enters the queue for a human agent to resolve the query or triggers another robotic automation process such as an address correction.


2. Amplify human intelligence. In these applications, AI is not replacing people, it is giving them the tools to increase their capabilities and perform their jobs more efficiently.

only 12 percentRetailers are using this approach to create more positive conversations with customers and drive long-term customer engagement. One example is to use AI to determine how and when to intervene at the exact right moment before customers complain. AI could detect an unusually high bill, relative to that customer’s history and initiate a proactive communication.

Another approach in retail is to give tools directly to the customers, often in the form of apps that give them greater control. Theme park apps enabled with AI allow customers to spend less time in “traffic.” Also, many retail electricity providers are giving apps to customers that show the electricity consumption rate by appliance and offer suggestions as to how they reduce both consumption and cost.

Some companies are also building expert assistants with advanced search and question-answering capabilities to help researchers better service customers. These have proven particularly valuable in financial services, where research analysts must perform company and industry analysis, forecasting and valuations—all of which help make informed recommendations about a customer’s financial portfolio.

Analysts must collect and review many disparate sources of complex data and have the technical and market expertise to analyze and interpret it intelligently. This means they need to be an effective forecaster, modeler, valuation expert and risk manager. Expert assistant and intelligent search applications are helping with this. They can understand spoken or written natural language questions and deliver answers relevant to financial services or other domains as required. They’re able to search and pinpoint information based on semantic, contextual and proximity searches, as well as disambiguate, filter and group information. Crucially, they can also improve the quality of responses over time based on usage and feedback.

This means they can deliver significant productivity gains along with greater accuracy of information, improved customer satisfaction and, ultimately, greater competitive advantage. In this way, some firms are increasing research analyst productivity by enabling each analyst to cover significantly more equities than was traditionally possible, while increasing the quality of research.


3. Automate insight generation. AI can make the analysis of Big Data far more efficient so that companies can extract more value from their information, faster. As volumes increase, firms are after more insights and they often need it instantly throughout the day. The challenge is they must sift through a lot of information to identify what is useful and then ask the right questions. AI can tackle huge volumes of fast-moving data, and wrap around these analytics so it identifies those insights and finds the best ways to react.

smaller nimble companiesRetailers, for example, are constantly trying to understand more about their customers to form better quality interactions and drive revenue. AI is helping them move beyond traditional approaches and achieve levels of insights that never existed before. They are capturing and connecting all the data from each interaction across multiple channels, including website, social media and in-store, to create a truly 360-degree view of each person. With AI technology, retailers can combine social behavior (what they say) with consumer behavior (what they do), plus casual intelligence to look at what all of it means. They can then see individual data-driven journeys and use those on a wider scale to create better engagement.

For example, the traditional approach might envision a “day in the life” of a theoretical suburban mom based on perceived personas. This might involve getting children ready for school, shopping during lunch and family time in the evening. However, with unique, granular and time-stamped data, retailers can see actual behavior. For example, her children watched Dora, she shopped on Peapod during lunch and visited WebMD in the evening. AI intelligence can then layer on what they did, when not on the retailer’s website, and begin to build a more complete and useful picture. This can be leveraged both for broader strategic decisions, but more importantly, it can now be used to engage users in the most impactful micro-moment of their customer journey.

A major strength of AI is that it helps retail companies gain a more accurate business outlook by processing a variety of data—from in-store traffic to social-media sentiment, stock prices and consumer behavior. It allows companies to blend quantitative data with telling qualitative data to provide a complete picture. This holistic view of the business is critical for day-to-day operations, expansion planning, testing new products and services, and more.


Uncovering New Opportunities

Although there are some tangible successes, the AI revolution is still in its infancy and the race is on to see who can deliver the next innovation. While many will develop AI within each of the three categories of application, some companies are taking it a step further by combining different elements from each. In doing so, they are able to create new business lines and achieve completely new commercial gains that were not feasible before AI.

Earnest, a tech-enabled loan company, is a prime example. Rather than rely on predetermined factors such as credit scores for assessing loan-worthiness for a handful of predefined loan packages, it collects a wider set of data for each customer and then computes a completely customized loan package. By integrating multiple different AI approaches, the company has brought a totally new, personalized concept to the loan market, one that was not economically feasible without the use of AI.

More applications will emerge quickly, as there are still many unexplored opportunities. According to IDC, the AI market could grow from $8 billion in 2016 to over $47 billion in 2020. That is a huge leap and it will be those early adopters that are in the best position to claim the market lead. Now is the time to act and not just join the revolution but help to drive it forward.

Rashed HaqRashed Haq, vice president of analytics and artificial intelligence at Sapient Consulting, is responsible for helping companies transform and create sustained competitive advantage through innovative capabilities, processes and solutions in their business operations. His experience covers artificial intelligence and advanced analytics, including simulation and optimization methods for deterministic and stochastic problems, machine learning and causal reasoning. Prior to joining Sapient, he conducted research at the Institute for Theoretical Science and Los Alamos National Lab.