The sensational way the media often portrays artificial intelligence (AI) can make it seem intimidating. But to grow in today’s environment, businesses need AI to help identify patterns and problems within their masses of data. In this helpful text, analytics expert Asha Saxena takes the stress out of understanding the practical role that big data and AI should play in your business operations. She reminds you that you probably already use AI in your daily life, even if you’re unaware of it, and she explains how business leaders can adapt AI to their specific organizational needs.
- Netflix and Starbucks were big data and artificial intelligence (AI) pioneers.
- Big data fuels AI.
- Data and AI can reveal new sources of value within any industry.
- AI poses serious ethical issues.
- Assess your business before trying to implement a big data and AI strategy.
- Ask: Is your organization ready for a data-driven strategy?
- When implementing big data and AI, focus on the areas where you can generate the most value.
- To use AI effectively, you need the right data team.
- As a business leader, be mindful of the emergence of “Web3.”
Netflix and Starbucks were big data and artificial intelligence (AI) pioneers.
In the 1990s, Blockbuster Video was the dominant, go-to company to rent the latest video or DVD to watch on a Saturday night. But by 2013, the company imploded, and Netflix raced ahead. Blockbuster failed to recognize the internet’s transformative potential, while Netflix’s founders embraced change and the possibilities of the internet.
Recognizing that Netflix’s future lay in streaming video, the company developed an increasingly precise algorithm that recommended movies to subscribers and registered their preferences. Netflix then began producing original content which became highly successful.
“Without Netflix’s focus on big data and AI, no one would easily create original shows like ‘Orange is the New Black’ or ‘Stranger Things’.”
By 2020, Netflix upended the traditional movie and television business as it became a gigantic enterprise with more than $2 billion in profit. It succeeded, in part, because its leaders were willing to break traditional business models. They exploited and refined their data, focused on its inherent value, and based their business decisions on results provided by data analysis and AI.
Netflix is not the only company that leveraged data and AI to expand dramatically. Starbucks went from a single Seattle coffee shop to a ubiquitous franchise by using data and AI to improve its customers’ experience in various ways, including online ordering and a “loyalty” app. Both firms use data to understand their customers better, so they can deliver what their buyers want.
To identify the right data, be clear about your objectives. Identify the customer issues you want to fix. Make sure your data is relevant to your larger business goals, so you can leverage it effectively. Finally, amass a mix of historic and predictive data, that is, information about what your customers have done so far plus insights about what they might do in the future.
Big data fuels AI.
Artificial intelligence, or AI, consists of several continuously evolving technologies. Broadly speaking, AI isn’t the Hollywood version of artificial intelligence which posits that robots will be able to do everything that people can do, and more. Rather, AI is a computational procedure that can mirror some fundamental features of human intelligence, such as reasoning and analyzing information.
“AI is capable of extraordinary feats. It can free humans from repetitive tasks and exponentially increase the value that businesses can deliver.”
You can divide AI into three strata. Straight AI mirrors the capabilities of human intelligence. “Machine learning” consists of algorithms that enable learning from data. “Deep learning” is machine learning that applies the structure of neural nets to staggering amounts of data.
Though it doesn’t necessarily require AI, “data analytics” also comes in three forms. “Descriptive analytics” offers an account of what information is in a set of data and its history. “Diagnostic analytics” provides an analysis of what is in a data set and why. “Predictive analytics, the most important form for business, assesses what might happen in the future based on today’s data. Predictive analytics allow people to forecast, at some level of probability, what customers may want or do in the future.
Big data is crucial for AI, since it provides a vast amount of information for AI to learn from and analyze. Big data consists of “three Vs”: “Volume” is the sheer amount of data, whether gigabytes or even petabytes. “Variety” denotes the diversity and complexity of the data. And, “velocity” is the speed with which your company can possibly process the data. When AI with machine learning analyzes your company’s big data, you no longer need to make significant business decisions in the dark or based on any individual’s limited purview or intuition. Instead, you can base your decisions on objective information.
Data and AI can reveal new sources of value within any industry.
The “AI Factor” is the value that resides in real or potential data in almost any industry. Any organization can benefit, financially or otherwise, from uncovering this value by using AI to leverage data in innovative ways. The most successful “multiplier” businesses will be the ones that are open to changing their business models.
“The good news is that AI and big data are more accessible than ever. Although not without difficulty and risk, they are the keys to extraordinary success.”
Take Domino’s Pizza. Though it was successful in the 1990s, Domino’s suffered during the 2008 recession. It responded by using smart devices in its marketing. In 2012, Domino’s introduced a Facebook page called “Think Oven” that solicited customers’ ideas and preferences on various subjects, including its menu and even the look of its employees’ uniforms.
The initiative went well, so Domino’s increasingly used data and digital technologies, which became the core of its business model. Customers could order pizza on Slack, Facebook Messenger, Twitter and Domino’s mobile app. It also set up AI virtual assistants to help customers place their orders. By 2017, Domino’s had become the world’s largest pizza chain.
AI poses serious ethical issues.
Big data and AI can contribute significantly to an organization’s success. However, unscrupulous people also can use them in manipulative, unethical ways. Unethical data usage can have an impact not only on individual organizations but, potentially, on an entire country’s political processes, communities and individuals. Officials have called out major social media platforms like Facebook for their questionable use of customer data. The company may ultimately face government investigation, regulation and prosecution. The unethical use of big data and AI has a negative impact on society – and it’s bad for business.
“[AI and big data] can also have tremendous effects – positive and negative – on people and communities, and on the planet where you do business.”
Ethical AI is transparent. The way an AI system works should be clear. It should operate in a manner consistent with moral values, such as dignity, human rights and equal treatment for everyone. Designers should not create AI systems for destructive purposes, and the people who use AI should take responsibility for how their systems operate. Finally, those who generate the data AI uses should maintain control of it. Simply establishing an ethical framework for big data and AI may not be enough. Jurisdictions may need to set legal regulations for using AI and big data, especially when it comes to meeting international standards and protecting personal privacy.
Assess your business before trying to implement a big data and AI strategy.
Big data and AI have immense transformative potential, but if you’re a relative beginner, you need to learn how to start creating a strategy for using these tools and making the most of their potential. Apart from convincing nervous colleagues that the strategic business use of AI and big data is a good idea, evaluate your organization’s capacity for growth and innovation. As you figure out how to use a data-driven approach effectively, customize your strategy to handle your firm’s unique challenges.
“Doing a growth assessment also forces us to look at the type of company we now are and, more importantly, what we want to become.”
First, determine if your company is open to risk-taking and innovation, because it may need to give up some current business approaches. Second, assess whether it has the capacity for future growth in its current sector. Organizations that are open to dynamic innovation tend to have executives who push new approaches; strategies that give employees incentives for innovating; strong, accurate metrics that predict when innovation is likely to work; and policies that encourage open communication and reduce anxiety over risk-taking. Organizations often underestimate their capacity for growth. Often, people just don’t pursue enough opportunities.
Ask: Is your organization ready for a data-driven strategy?
Once you determine that your firm is open to innovation and growth and can benefit from big data and AI, make sure it’s ready to adopt an AI and data-driven strategy. To be ready for that transition on a managerial level, your company needs executives who are willing to become invested in its AI initiative and technology leaders who are aligned with the overall firm and its goals. To be “data-ready” for this transition, your company needs access to significant “structured” data – such as names, phone numbers and email addresses – and “unstructured” data – like images.
“The good news is that most companies have (or at least have access to) the data they need, whether they know it or not.”
An organization can access data in numerous ways. For instance, it can set up “active data collecting” to draw information from surveys, questionnaires, focus groups and interviews. Social media’s vast amounts of text, images and videos are another source of data. Your organization’s “passive or permissions-based data collecting” can include customer-service communications and e-commerce. This data provides information on user preferences, interests and purchasing decisions. Public records are another data source.
Finally, by 2024, a more recent phenomenon called “synthetic data” will encompass more than half the data that AI and machine learning use. An algorithm that creates new data by modeling the probability distribution of an existing real data set generates synthetic data.
When implementing big data and AI, focus on the areas where you can generate the most value.
Once your organization is data-ready, identify the activities where AI and data collecting will generate the most value.Don’t focus on multiple areas or products at once – at the beginning, one or two are more than enough. Once you’ve chosen a focus, you’re on your way to becoming a “multiplier”: a company that grows exponentially.
“By demonstrating your first notable success in the power zone, you will draw others’ attention to the potential of data to multiply business performance.”
AI and big data can have their most significant impact when applied to your organization’s biggest unsolved problems. When considering which problems fit the bill, think about how much specific issues are costing you overall and how they could make you miss out on meaningful future possibilities.
Data and AI may not be able to help you identify your most significant unsolved problems. Leaders may have to rely on their intuition or “gut instinct” and then move on to assessing their data. Still, be mindful of your circumstances. Companies such as Netflix don’t pursue their data-driven strategy in a vacuum. They continually consider external issues, such as market conditions, inflation, interest rates and government policies.
As you move forward, establish clear goals, ranging from small and incremental objectives, like controlling costs and managing prices, to large and substantive targets, such as expanding your customer base and developing new products. Be aware that analysis of your data may disprove your operating assumptions and – given that AI works with a vast amount of data – it may lead you to places you never thought you’d go.
To use AI effectively, you need the right data team.
At this point, you’ve evaluated your organization, established its data-readiness, and set priorities and goals. To implement your big data and AI strategy, you’ll also need the right staff, including an effective data team. That team should include a senior advocate, an engineer, a data scientist, an operations engineer to oversee implementation, and a business specialist to make sure the technical people’s work aligns with the organization’s business needs and goals. The data team needs to be familiar with the appropriate approaches to data and able to focus on the project’s goals.
“If a project’s measured performance matches or exceeds that of its predicted business value, then it will be the cause of real business growth.”
First, the data team must establish your data’s value and determine whether it’s capable of supporting the organization’s objectives. Second, the team must figure out how to collect still more data. Ideally, the new data will work with machine learning, thus allowing more precise and accurate AI assessments and decisions. Finally, the team must determine a way to measure the results of its work with data and AI. After all, you can’t take data’s value and impact for granted. Your project’s results must be measurable.
When developing precise measurements for a data-driven initiative, be wary of assumptions and cognitive biases. Mistakenly conflating correlation with causation is easy, as is succumbing to the temptation to seek and accept evidence that supports your existing assumptions even if it doesn’t pass a quality evaluation. To mitigate such cognitive errors, create a peer review system, set well-articulated ground rules and try to establish a corporate culture that focuses on each project’s larger goals.
As a business leader, be mindful of the emergence of “Web3.”
The prominence and development of big data and AI have led to the emergence of “Web3.0” or “Web3,” in which data is autonomous and decentralized. This has given rise to a host of new technologies, including blockchain, which is based on an encrypted ledger. A blockchain leaves data independent and in its user’s control, and it ensures privacy. The blockchain is crucial to the emergence of cryptocurrency. Though the blockchain has received a great deal of negative press, it has many other potential uses. Thus, leaders who are developing future-oriented big data and AI strategies should be mindful of Web3.0’s development, since most of what they are doing now is still rooted in Web 2.0.
About the Author
Asha Saxena, an adjunct professor of health policy and management at Columbia University, teaches healthcare consulting, entrepreneurship, big data and data analytics. She was CEO of Aculyst, a healthcare analytics firm, and president and CEO of Future Technologies, an international data management and analytics company.
This document is restricted to personal use only.