At any given time, a business leader is in pursuit of numerous goals– achieving growth targets, entering new markets, beating the competition, ensuring customer satisfaction, reducing costs, embracing new technologies and maintaining organizational morale, to name just a few! To achieve any of these goals, leaders need timely and relevant information rich with insights. The digitization wave has led to an explosion of data, from both within and outside the organization. This information is a treasure-trove waiting to be harnessed. Any business serious towards achieving its goals can ignore the latent power of digital data only at their own peril.
For a long-time now businesses have been hampered by lack of information, sub-optimal decision-making and a general lack of systemic business intelligence. The traditional rigid pyramidal organizational structure with limited modes of communication and data measurement are increasingly becoming a hindrance rather than enablers. Recent developments in technology have unearthed new ways of solving a plethora of old business problems. Big data and artificial intelligence have exploded on the global corporate landscape with potential future impact no serious stake-holder can ignore.
Business in the age of disruption
The internet emerged in 1990s and forever changed how businesses worked. In a single swoop all information exchange became potentially digital and instantaneous. The mobile revolution which followed made communication all-pervasive & universal. The last 2 decades have seen more technological breakthroughs than any other generation in human history. Cloud computing transformed the way businesses stored data, big data revolutionized pattern-recognition, genetic engineering transformed the medical industry and e-commerce forever changed the global retail landscape. While disrupting technologies have improved the quality of our lives, for companies in competitive markets, it’s always a struggle to adapt, evolve and survive. New technologies mean newer processes, investments and more skill up-gradation requirements to stay relevant and ahead of the competition.
Only a decade ago, businesses were content with traditional technologies which seem archaic by today’s standards. Gone are the days when data entry and analysis was done on MS Excel and data visualization was done predominantly on power-point presentation while accounting was done on simple tools like Tally. As markets have become more fluid and dynamic, organizations have grown flatter and larger in response. Reliance on old ways of doing business have proven increasingly inadequate. The age of disruption demands a disruptive solution to its problems!
Era of data explosion
With the growth of information technology and increased digitalization of business processes, data of unprecedented volume, variety, and veracity is being generated as free by-product of business operations. This data set is so big and complex that it needs an altogether new set of technologies to process it. This data & its associated processing technologies came to be known as Big Data. The graph below surmises the magnitude of the data explosion which businesses across the world have experienced till date combined with forecasted data explosion foreseen by experts in the coming decade.
Note: 1 Zeta byte = 10^27 bytes (or 1 billion billion Giga bytes)
The accelerating pace of data generation today is mind boggling. It is estimated that 90% of all data in the world has been generated over the last two years alone. Today, businesses generate a stupendous 2.5 quintillion bytes of data per day and the rate is growing.
The need to augment traditional business intelligence
The problems of rising decisional complexity arising out of new technologies and larger organizational size is not new. Throughout the 20th century businesses have grappled with these problems and have dealt with them in several ways. A popular way has been to hire management graduates from premium universities to fill leadership roles. This is based on faith on managerial education’s capability to handle ever-evolving business problems. Another popular way is to hire services of strategy consulting firms. Yet another way is frequent skill and knowledge upgradation of workforce by trainings. But in wake of recent data explosion, most businesses found that these traditional business intelligence modes need to be augmented by continual data-analytics system which automatedly churns out data-insights. The case to augment traditional business intelligence modes with a complementary data-system is made below by highlighting varying, hitherto unrealized accruing benefits:
- Speed: An urgent task requires immediate solutions. Consulting, inhouse training or lateral hiring of experts would require help of systemic data architecture with insights into major business metrics in order to take relevant decisions in times of crisis.
- Accuracy: Computers are better at repetitive tasks than humans. For large-scale repetitive iterative tasks, an automated system could be more accurate in predictions while leaving the value-based decisions for their human counter-parts. As cross-sectoral & innovative thinking is a uniquely human trait, a combination of human and computer systems is most conducive for engendering holistic accuracy.
- Imparting scalability: Ever changing and voluminous machine generated data require a more automated and systemic approach. But how to fit data-generated insights into the macro-economic framework is best done by organizational leaders or strategy consultants. The complementarity of functions is obvious.
- Cost-optimization: Hourly billing rates for a typical consultant vary from $500-$3000/hour. Similarly, investments into new data-architectures also have a base threshold cost. These aren’t usually financially viable for many cash-starved innovative startups. An alternative approach can be pooling of resources by multiple startups for a common pool of consultants and data architecture solutions.
Sole reliance on traditional software systems and business consulting services is no longer advisable for long-term sustainability. These need to be augmented by data-systems to better embrace newer possibilities. A testament to the adage “Necessity is the mother of all invention”, big data & analytics answers the call for of most of these problems.
Leveraging Big data
According to PwC’s CEO Survey, the top 3 business priorities for CEOs are innovation, grooming human capital and developing digital & technology capabilities. It is clear that most leader’s want to actively invest in enhancing business intelligence by investing in digital technology, personnel training and incorporating new systems.
Armed with sophisticated algorithms capable of vetting through vast data-sets, big-data analytics offers business leaders a bottoms-up option to deal with their data problems. As the industry matured, big data is increasingly becoming available for smaller firms as well. Customized dashboards, automated reports and live bot customer support are all a reality today. But hardly anyone expected that all this immense development would soon seem like a mere tip of the iceberg. It’s only in the last few years that experts & world at large have realized that the real revolution which is beckoning is the incoming wave of Artificial Intelligence (AI)!
The rise of Artificial Intelligence
Artificial Intelligence (AI) is the natural progression of the field of big data and analytics. AI systems aim to perform tasks normally requiring human intelligence, e.g. pattern recognition & decision making while automatically improving its own learning abilities without active external intervention. Unlike the human brain, a computer does not have an upper limit on size and hence there is no upper limit on its memory, potential learning or intelligence capabilities. Many experts forecast that by 2045 AI systems will surpass average human-level intelligence (also called Artificial General Intelligence-AGI).
AI is to big data what the development of steam engine was for human terrestrial transport. While big data solutions available today have be to custom-developed for each type of data architecture, AI promises to provide seamlessly integrated, customized, automated business intelligence solutions across platforms like ERP, CRM, financial or human resource databases. Where-ever there is data and possibility of deciphering insights, there is scope for AI deployment.
Rising demand for AI capabilities
The prime driving force behind adopting AI in businesses is competitor advantage. It is changing the way businesses operate. A survey conducted by MIT Sloan about why companies are adopting AI shed light on a typical business leader’s prime motivations for adopting AI.
A survey by Tech Pro Research estimates that around 24% of businesses currently are implementing or plan to implement AI projects. Leading sectors in AI adoption are automotive, health, financial services and software automation.
Investments in AI
PwC estimates that AI could add $15.7 trillion to global GDP by 2030. According to IDC AI Spending Guide, global expenditure on “cognitive and artificial intelligence (AI) systems will reach $19.1 billion in 2018, an increase of 54.2% over the amount spent in 2017” (IDC, 2018). It further predicts a CAGR of 46.2% over the 2016-2021 forecast period.
- Investments in AI startups: Research firm CB insights released its 2018’s “AI 100” report highlighting that 100 startups have raised over $11.7 billion in aggregate funding across 367 deals. China’s ByteDance which uses AI for personalized news recommendations tops the list with $ 3.1 bn in funding. 76% of the startups were from the USA reflecting the continuing dominance of the Silicon Valley in tech.
- Non-Tech Fortune-500 companies playing catch-up: A report by CB insights shows that for the first time, non-tech fortune-500 companies are investing more in technology in an effort to survive. In 2017, 51% of all investments into non-tech companies were from these corporations- Goldman Sachs invested in Uber and Spotify while Morgan Stanley invested heavily in Flipkart, Dropbox and Domo.
- Sectors betting heavily on AI: Retail, telecommunications, banking & media are the sectors investing heavily in AI followed closely by automobile, public safety, pharmaceutical research and product recommendations. (IDC, 2018).
- Corporate AI R&D spend: A McKinsey 2017 report estimates that online firms like Google and Baidu had around $20-30 billion invested on AI in 2016 with 90 percent of this spent on R&D and deployment, and 10 percent on AI acquisitions.
- VC & PE investments in AI: Venture capitalists invested $4-5 billion in AI in 2016 while Private Equity (PE) firms invested $1-3 billion during the same period.
How AI is transforming Business Intelligence
Although the possibilities are limitless, there are some AI applications which are ubiquitous in their potential to transform almost any industry. AI systems are being engaged in intelligent interactions with role-based dashboards. Apart from decision making, AI is being used for improving automating customer interactions, data mining, and improving the recruitment process. In every sphere of business activity AI is providing customized, scalable, flexible, fast and accurate solutions.
- Decision-making: A sophisticated AI system can run more than 200,000 business case scenarios within hours or even minutes compared to just a handful by human analyst over several days. The result is more informed, faster and better decisions resulting in revenue gain, cost reduction and gaining a competitive edge.
- Finance: AI algorithms are helping traders and hedge fund managers predict the trend in stock prices based on a holistic understanding of core company fundamentals, market trends as well as probabilistic calculations of impact by external factors such as pending regulations in local economies. 2017 saw AI Powered Equity, an exchange-traded fund that uses AI to select stocks it holds, as one of the most successful funds of the year.
- Marketing: AI tools in providers like Zoho Campaigns can direct email campaigns to the large audience by delivering the emails to each recipient at calculated times with maximum probability of opening by analyzing each individual’s historic email opening pattern. Not only does this obviate hiring additional staff, it also helps maximizing the ROI per each email campaign.
- Operations: AI in quality testing processes have increased defected detection rates by 90% making it easier for more and more organizations to reach the ever elusive Six-Sigma club.
- Procurement: Having a real-time view of inventory enabled by a dashboard system (big data application by systems such as Tableau) and forecasting of demand based on predictive analytics, the job of a manager becomes simpler. He/She knows how much to order & when while minimizing the risk of losing potentially big future purchase orders.
- Design: Major automobile players like BMW are increasingly using AI in designing cars with new features such as self-driving cars. BMW recently announced its partnership with Intel to use computer vision technology in its future designs. The aim is to enable cars to see just as humans do. Other companies like Tesla, Uber and google are close on the heels to develop the same capabilities in their fleet.
It is not surprising that most progressive businesses, organizations and leaders have their eye on utilizing the benefits of AI.
Case Study- Tesla: Driverless cars of tomorrow
Tesla has used AI and Big Data in a very effective way to gain a significant competitive advantage. Research by McKinsey & Co estimates that market for vehicle-gathered data will be worth $750 billion by 2030. Foreseeing the potential, Tesla has institutionalized its fleet’s holistic data capture processes and integrated it to an AI. It crowdsources data from all of its vehicles as well as drivers via sensors both within and outside the cars capturing information like driver hand placement, road congestion, traffic speed and location of hazards along routes.
The Tesla AI operates at 3 levels of learning:
- Machine learning in the cloud educates the entire fleet on variables like ideal routes, congestion data and hazard location (scalable mass learning)
- Networked intelligence is tapped by forming a network of Tesla vehicles in vicinity to share local information & insights (decentralized & customized actions/solutions)
- Intelligent computing at individual car level decides what action the car needs to take right now (speed)
In the future, plans are to integrate Tesla networks with cars from other automobile manufacturers along with other systems like traffic cameras, road sensors and mobile phone networks. The AI-driven partnership between Tesla and hardware manufacturer Nvidia is based on unsupervised learning model of machine learning. The potential for Tesla to exponentially improve its business intelligence using AI will bring it closer to its goal of providing comfortable and safe driver-less navigation and its ultimate goal of capturing the automobile mass-market.
Pharma Case study: Curing cancer with AI
The power of AI is profoundly felt when its used in saving lives. A study at the University of Heidelberg, Germany with over 58 international dermatologists showed that a deep learning AI variant called convolutional neural network (CNN) could better identify and diagnose malignant moles (skin cancer) than licensed experienced dermatologists.
BERG, a Boston-based pharma start-up is using AI-driven data analytics for drug discovery. They firmly believe that this unique combination of maths, chemistry and biology is key to discovering the correct drugs for complex diseases.
Apart from genetic information, a single cell can produce over 14 trillion data points. The BERG’s AI processes huge amount of biological data to reveal a cells journey from good health to cancer and allows informed hypothesis testing for efficient drug formulations and narrowing on choices for drug trials.
Utilizing AI scenario testing, BERG team discovered how cancer cells thrived within pancreatic cells by de-activating the mitochondria. By reactivating the mitochondria, BERG’s drug BPM 31510 essentially turns cancer cells back into normal cells. The team believes they can use this model for other forms of cancer treatment as well (NES).
AI adoption guidelines for newcomers
As business decision-makers realize the importance to enter the AI domain, the immediate challenge they face is whether their organization is ready to imbibe AI into their organizational fabric. However, not all organizations are ready to imbibe AI just yet. There do exist certain barriers to developing internal AI capabilities.
Organizations which haven’t adopted AI yet need not be gloomy however. Although building an AI infrastructure takes time and investment, it still is a very feasible path. By playing the cards right new businesses can enter the mainstream AI landscape within 1-2 years and gain a strategic advantage over their competition. Since not all companies are as adept at adopting to change, there is a clear early mover advantage.
Below are main milestones on path to developing internal AI capabilities:
- Beefing up digital capabilities: AI & big data are pattern-recognition algorithms built upon the basic building block of data. For an organization to introduce AI, there has to be a legacy minimal data volume and diversity (5 levels of data). It’s much like raw material to construct a building. This is a challenge for companies who haven’t already bought their systems online. Next steps would involve time-taking initiatives of buying computing power, brining systems online, training the work-force and then opening up to AI incorporation.
- Investment: There is a threshold prior investment required in terms of computing power, servers, recruiting experts and training. This acts as a hindrance for small companies or early-stage startups who have significant cash-flow issues. Therefore, most organizations who have reached a threshold size & cash reserves are in the best position to enter the AI landscape.
- Continual evolution: Adopting AI is not a one-time thing. Once adopted, the system evolves and necessitates constant upgradation of systems, processes and technology. The result is that for organizations to adopt AI, they have to embrace it completely. This includes initiatives such as constantly retraining employees on new processes, redefining processes. A change-embracing organizational culture is a pre-requisite for successfully harvesting any potential AI benefits.
- Active consumerism: It’s imperative that stakeholders involved in initial setup of AI systems in the organizations should know how the systems operate. Rather than being mere consumers, they should strive to be active participants and learners in an ever-evolving AI ecosystem. This will help organizations initiate changes in an AI system externally in the future if required.
AI-powered business intelligence addresses most of decisional-complexity problems which traditional business intelligence models were seeking to fill. It’s a fast, accurate, scalable and cost-effective solution system for decision-making. In fact, AI goes beyond generic pattern recognition and is even helping domain experts do their jobs better. Empirical evidence strongly indicates that AI-enabled companies are performing far better in terms of growth and sustainability compared to those who are yet to take the AI-plunge. The writing on the wall is clear: AI is vital and significant part of the future. Fortune favors the brave. Businesses which take the lead today have most to gain!