How Business Can Leverage AI

 

“The core currency of any business will be the ability to convert data into AI.”

Satya Nadella, CEO, Microsoft

 

“Just as electricity transformed everything almost 100 years ago, today I actually have a hard time thinking about an industry that I don’t think AI will transform in the next several years.”

Andrew Ng, cofounder, Coursera

 

Let’s explore how the powerful capabilities of AI, can be applied to business. Andrew Ng has compared the current state of AI to the discovery of electricity. This remarkably powerful general purpose technology has now entered an era of growth and implementation. In the same way that progress in electricity revolutionized the business world in the 1800s, improvements to AI in the 2000s are propelling a myriad new business opportunities. Amazon CEO Jeff Bezos, the wealthiest man in the world, has said we are in a golden age of AI that can provide “an enabling layer” for “every business.” As Bezos says, AI has been central to the success of

Amazon, optimizing a multitude of areas:

I would say a lot of the value that we’re getting from machine learning is actually happening kind of beneath the surface. It is things like improved search results, improved product recommendations for customers, improved forecasting for inventory management, and literally hundreds of other things beneath the surface.

But AI is not the exclusive preserve of Amazon and the other tech giants. Every business can be enhanced by AI. This includes large multi-national corporations, mid-sized companies and even SMEs. AI is within reach of all of us. Open-source code and bright people are all businesses need to make massive progress with AI. The opportunities are endless.

 

REIMAGINING BUSINESS WITH AI

Where do you start? The key thing to consider: AI offers unlimited opportunities for smart, effective, business solutions. There is not a fixed set of 20, 50 or 1,000 solutions. There are as many opportunities for applying AI to business as your imagination and innovative spirit permit. You need to reimagine business processes. Consider that applying AI to a business process could make a positive impact. New applications for AI in your business will create a new solution that only you are utilizing, giving you a healthy competitive advantage. Custom AI systems are a strategic asset that will increase the overall value of your business.

The first phase is focused on comprehension of the technology. This includes: (1) technical knowledge of how AI works and what it can do, and (2) market knowledge of how the technology will impact business models and evolve over time. This book aims to answer many of these questions.

 

ALIGNMENT WITH AI OBJECTIVES AND AMBITION

In the book The Executive Guide to Artificial Intelligence, Andrew Bur-gess argues that AI implementation should involve alignment with your strategic objectives. These may include:

1. Generate revenue
2. Reduce costs
3. Manage risks
4. Improve customer experience

You then need to consider your AI ambition. You might want to:

• Just tick the AI box to claim you have AI in your marketing materials
• Improve processes
• Transform a function
• Develop a whole new product, service or business
• Shift your business to become an “AI as a service” business. This links back to your strategic objectives, since you should be open to revising your business strategy based on your AI ambition.

 

BE STRATEGIC

My team and I at Critical Future don’t recommend implementing AI just to tick a box to claim you have AI in your marketing materials. This would result in a shallow implementation of AI without any long-term benefit or strategic value. We recommend, instead, developing AI systems that sup-port your business strategy.

Cynthia A. Montgomery, noted professor at Harvard Business School, explains: “A great (business) strategy is more than an aspiration, more than a dream: It’s a system of value creation, a set of mutually reinforcing parts.” ¹ Value creation involves these two outer limits: customers’ willingness to pay (essentially, customers’ satisfaction with a good or service) and suppliers’ willingness to supply (essentially, their opportunity cost— the lowest price at which they would be willing to sell to a particular firm).

When a company drives a wider wedge between these lines, by expanding the total value created, its existence matters in an industry. At this point, the company is much more likely to be able to claim some of the value for itself, meaning it is able to increase its own profitability without making its partners in trade less well off. A business creates value by driv-ing the widest wedge possible between the satisfaction of customers and the all-in costs of suppliers:

Figure 9. Value Creation, by economists Adam Brandenburger, Barry Nalebuff and Harborne Stuart

 

This is where your AI system development should be focused—on supporting you to drive the widest wedge to create value. Ask yourself what matters most about your company. If your organization were to disappear tomorrow, who would miss it most? The AI systems you develop should make your company matter more to the people who care most about it.

 

ASSESS OPPORTUNITIES

You can then conduct an AI opportunity assessment to ascertain where AI can be applied to your business. You can use AI to solve external sec-tor wide challenges. For example, at Critical Future we have developed AI systems to predict property values, predict cancer, predict stock prices and predict commodities. To identify external opportunities, simply think what is the most valuable thing you can predict in your sector? Think through where to find the data, and come and talk us at Critical Future and we will build an AI System leveraging deep learning to predict it for you. I have no doubt whatever statistical models are currently being used to make these predictions we can beat them with AI.

You can also use AI to solve internal challenges facing your specific business. To help with this, we provide a framework here of specific opportunities to apply AI to every business function.

While all these opportunities for AI enhancement will support your business, you can prioritize them based on the strategic value they offer. The AI opportunities will fit some types of businesses better than others, within different contexts. For example, e commerce companies may find that using AI to enhance logistics provides a competitive advantage. This can increase customers’ willingness to pay and suppliers’ willingness to provide. B2B business-services companies, such as consultants, lawyers, accountants and recruiters, may find that the best value lies is increasing sales through AI. This can enable sales teams to be more effective, win more customers and increase lifetime value. B2C consumer-facing companies like retailers may find the best opportunity for AI is in marketing, where it is possible to personalize the experience to every customer, ensuring each prospect receives the right message, in the right way, at the right time. Manufacturing businesses may find the best opportunity for AI is in operations, to increase productivity and lower fixed costs. But there are no hard and fast rules here; it is up to you to decide where AI can deliver the most value for you.

The opportunity assessment must be done on a case-by-case basis. It involves three key considerations:

1. Strategy: What is the opportunity for value creation? Will this AI system make my company matter more to the people who care about it the most?
2. Data: Is the required data available or obtainable?
3. Resources: Can we achieve this AI system either through buying ready-made AI software or through hiring consultants, contactors or full-time team members to deliver?

 

AI BUSINESS SOLUTIONS FRAMEWORK

Function	Challenges	Intelligent Solutions
Supply Chain	• Enhance visibility.	• Control software responds automatically to scale supply chains, in response to real or predicted demand.
	• Find weak links.	• Predictive intelligence anticipates demand based on previous demand fluctuations.
	• Match demand with production.	• Computerized suggestions help resource managers make decisions with all available information.
	• Remove bottlenecks.	• Heuristic algorithms produce rapidly convergent load-spreading solutions.
	• Scale supply.	• Machine learning, combined with IoT devices and intelligent monitors, automatically flag failing or slow links in supply.
	• Manage fluctuating demand.	• Natural language processing (NLP) for data cleansing and building data robustness
	• Schedule deliveries.
	• Manage suppliers more effectively.
Sales	• Discover new customers.	• Analytic systems explore existing customer graphs to find new customers.
	• Find cross-sell/upsell opportunities.	• Supervised (or reinforced) learning systems analyze existing e-commerce data to personalize user experience.
	• Prioritize customer types.	• Ranking systems prioritize customers most likely to respond to upselling.
	•Obtain customer insights.	• Intelligent sales lists order customers by most urgent.
	• Increase conversion rates.	• Automatic data analysis tools provide insights from raw customer data.
	• Automate admin processes and data management.	• Virtual sales agents can engage with customers as chatbots or conduct sales admin, such as automatically transcribing sales calls.
	• Increase sales force with digital employees.
Operations	• Predict maintenance and risks.	• Predictive maintenance systems estimate probability of failure and/or time to failure of production components.
	• Optimize production lines.	• Machine vision provides automated visual inspections for quality assurance.
	• Anticipate and schedule maintenance.	• Robotics automates production lines.
	• Quality Assurance	• Predict which processes, people, tools will deliver optimum productivity.
	• Improve productivity.	• Predictive asset utilization uses power when you need it and doesn’t when you don’t to preserve energy.
	• Preserve energy.	• Personnel systems learn from employee data and past performance to allocate the best available employee.
	• Schedule tasks.	• Use planning AI techniques to schedule dependent tasks in an optimal order.
	• Plan effectively.	• Accurately estimate project timelines with all available data based on previous projects.
	• Management insights and dashboards, live reporting and predictions
Marketing	• Build brand avatars.	• Rank or cluster existing customers according to interests, and tailor advertising to them.
	• Design campaigns.	• Identify information flow and marketing effect on test subjects with eye-tracking and other computer vision techniques.
	• Track campaign effects.	• Receive reports on campaign effect with automatic summarization of multiple market metrics.
	• Plan future marketing areas.	• Identify lucrative product areas for marketing with ranking systems.
	• Predict future trends.	• Capture product sentiment and “trending” concepts with natural language systems to monitor and predict market direction.
	• Target and personalize advertising.
	• Generate marketing content.
	• Listen to your audiences online.
IT	• Prevent cyberattacks.	• Predict and prevent maintenance requests with online predictive maintenance systems.
	• Triage automatically.	• Automatically assign incoming ticket priority and/or tags with natural language processing systems.
	• Create custom intelligent software.	• Ensure in-house developers are AI-literate to produce intelligent company tooling where required.
	• Carry out penetration testing.	• Constantly test deployments with intelligent tooling; alert administrators to security flaws and misconfigurations.
	• Identify and report network security breaches	• Monitor network traffic and topologies with alerting systems for unusual activity.
Human Resources	• Identify talent.	• Explore social/business networks such as LinkedIn and compile candidate lists with graph-processing AI.
	• Conduct Interviews.	• Filter, schedule interviews and even assess candidates automatically with chatbot/language processing systems.
	• Review performance.	• Highlight extremal performers in a large company with one-class supervised learning techniques.
	• Automate processes.	• Process internal communications with sentiment analysis tools: Keep an eye on the mind-set of your employees.
	• Gauge emotions.	• Use virtual agents to replace human operators to reduce admin.
	• Assess talent.	• Automate data capture and analytics processes.
Finance	• Report trends.	• Filter and refine voluminous financial reports with AI recommendation systems.
	• Financial analytics	• Predict and flag developing issues, such as unexpected losses, spiraling budgets or cash flow problems, with sequence learning.
	• Fraud detection	• Regression tools objectively apply learned numerical procedures, such as those found in risk management.
	• Credit risk	• Analyze and understand how account holders spend, invest and make financial decisions; customize advice for customers.
	• Reviewing contracts	• Index documents; extract information.
	• Procurement	• Virtual finance agents communicate with suppliers etc.
	• Robo-advice	• Robo-advisors give financial advice.