Machine customers have the potential to revolutionize commerce, but there’s a potential dark side
Simply put, machine customers are machines that can buy things on their own without human intervention.
For example: a smart refrigerator detects its running low on milk and based on your preferences and past orders, automatically places an order for your usual brand through an online grocery store.
Amazon Dash Replenishment is just one example of machine-driven purchasing. This Amazon service, integrated with Alexa, allows connected devices to reorder supplies or replacement parts. Once set up, customers can track supply levels within the Alexa app, receive notifications from
Alexa when supplies are running low or parts need replacement, and smart reorder from Amazon when needed.
Machine-driven purchasing may represent the biggest new growth opportunity of the decade.
According to Gartner, “By 2030, executives believe at least 25% of all consumer purchases and business replenishment requests will be substantially delegated to machines.”
Even if only a small percentage of the over 9.7 billion connected IoT devices becomes a customer, the economic impact could be staggering. However, this optimistic outlook presumes that consumers will jump at the chance to grant purchasing permissions to a wider range of potentially hackable devices in addition to their already vulnerable phones.
What are machine customers or “custbots”?
Machine customers are devices that use artificial intelligence and the Internet of Things (IoT) to make purchasing decisions on behalf of humans or themselves. A “custbot” is a shorthand term for machine customer, also known as an AI customer or Autonomous Buying Entity.
How do machine customers work?
Machine customers automate purchasing through integrations with payment systems and delivery mechanisms. They can make decisions about what to buy independently, based on criteria selected by the user, and past order patterns.
Their “choices” are fueled by data collected through sensors or existing, external systems. This data could include inventory levels, usage patterns, real-time prices, connected systems, databases, or even weather APIs, providing a broader context for informed decisions.
Machine customers analyze the data they collect to identify needs, compare options, and make the right decisions. Based on the data, pre-defined parameters, or learned models, the machine customer triggers a purchase.
What are some types of machine customers?
There are three basic types of machine customers. Self-serving, agent-based, and predictive.
Self-serving Machine
Self-serving machines buy for themselves, like smart appliances placing automatic online grocery orders through connected platforms.
Agent-based Machine
Agent-based machines act on behalf of humans, executing pre-approved purchases, like a fleet of company cars automatically refueling at designated stations with the best price.
Predictive Machine
Predictive machines anticipate future needs and order proactively based on data analysis, like factory machines autonomously bidding on energy during off-peak hours.
What are the three phases of machine customers’ evolution?
Gartner reports that machine customers will evolve through three phases: bound customers, adaptable customers, and finally, autonomous customers.
Bound Customers
In the first phase, machines will automatically perform limited functions as a “co-customer” on the owners’ behalf. People will set the rules and the machines will execute tasks within a specific ecosystem. Think about the example of Amazon Dash Replenishment from earlier in this article.
Adaptable Customers
In phase 2, people still set the rules for machines although AI tech can choose to act on behalf of a human with minimal intervention for select tasks. A good example of this phase of machine customers includes autonomous vehicle systems from Tesla or Google.
Autonomous Customers
In the last phase, these new economic actors have enough intelligence to act independently on behalf of humans with a high degree of discretion and own most of the process steps associated with a transaction. Machines in this phase will also be responsible for meeting their own maintenance and software update needs.
Emotionless Purchasing
Machine customers in each phase will make decisions differently from humans with significant commercial and operational impacts.
Machine customers can collect, process, and weigh data to make purchasing choices without being influenced by emotion. This focus on completing tasks efficiently will require companies to engage with AI customers in ways that differ from human consumers.
Machine customers are logic and rule based, so although the reasons for their choices will be more transparent, their problem-solving approach may not make sense to humans and have unintended consequences, especially when complicated algorithms are employed.
What are the benefits and advantages of machine customers?
Machine customers could increase efficiency and sustainability, stimulate economic growth, and improve access to goods and services for wider populations. However, the way in which machine customers are developed and implemented will influence whether they will ultimately help or hurt human society.
Let’s review the potential benefits of machine customers and the factors needed to enable ethical implementation of this technology:
Potential Benefits
- Increased efficiency and sustainability – AI customers can find bargains, negotiate deals, reduce costs, and minimize waste, optimizing resource management, and potentially leading to more sustainable practices, benefiting everyone.
- Economic growth and innovation – Machine customer technology could create new jobs in development, maintenance, and data analysis, stimulating economic growth and innovation.
- Improved access and affordability – By optimizing purchasing and logistics, machine customers could potentially make goods and services more accessible and affordable for wider populations.
- Streamlined logistics – Intelligent purchasing can ensure smooth delivery and resource allocation.
- Personalization and Customization – AI customers can tailor purchases to individual needs and preferences, enhancing user experience and satisfaction.
- Sustainability and Resource Conservation – Machine customers can prioritize eco-friendly options and optimize usage, leading to a more sustainable future.
- 24/7 Availability and Market Integration – Continuous operation facilitates real-time responses to market fluctuations and changing needs.
Enabling ethical implementation
To enable the ethical development and implementation of machine customers, several safeguards will be necessary. These safeguards include transparent development and open scrutiny, strong regulations and guidelines, and inclusive stakeholder involvement.
Machine customers’ algorithms should be open to scrutiny and address potential biases, as biases in algorithms can lead to discriminatory outcomes, unfairly disadvantaging groups based on factors like race, gender, or socioeconomic status.
Diverse voices, including workers, consumers, and communities, should be involved in shaping the future of this technology. If a small number of corporate players control biased algorithms, it will create unfair advantages and distort market competition.
Governments and international bodies need to establish clear regulations and ethical guidelines for the development and use of machine customers. Governments and businesses have a responsibility to protect consumers and their personal data.
Consumers have a right to understand how decisions are made, especially when algorithms affect their purchasing power. Transparency and accountability in the development and implementation process builds trust and accountability.
Machine customers collect and use personal data to make purchasing decisions. Privacy concerns are valid as data breaches and theft can have long-lasting negative impacts on consumers. Governmental and public scrutiny ensures responsible data handling and minimizes privacy risks.
What are some of the challenges and concerns associated with machine customers?
Machine customers have the potential for both good and harm. It’s important to remember that technology is a tool, and its impact depends on how we use it. Acknowledging potential risks allows us to take action to ensure responsible development.
Job displacement
Implementation of machine customers comes with the risk of potential job displacement as automation impacts jobs in purchasing, logistics, and related sectors.
Concentration of power
As machine customers automate purchasing decisions, power is centralized in the hands of the corporations and developers who design and control the algorithms. This could lead to further consolidation of wealth and influence. Dependence on machine customers could potentially exacerbate existing inequalities.
Potential for exploitation
If corporations use machine customers to prioritize profit over all else, the exploitation of resources and people in the pursuit of cheaper materials and labor will perpetuate existing inequalities and environmental harm.
Bias & data privacy concerns
Bias in algorithms and data privacy issues could worsen existing social and economic injustices, further marginalizing vulnerable groups.
Overconsumption and environmental impact
Prioritizing efficiency without ethical considerations could lead to increased consumption and negative environmental consequences.
Embracing the Machine Customer Economy
Machine-driven purchasing may represent the biggest new growth opportunity of the decade. Even if only a small percentage of the over 9.7 billion connected IoT devices becomes a customer, the economic impact could be staggering.
Businesses excited by this new technology can prepare by exploring the potential impact of machine customers on their specific industry and market and developing robust data quality, security, and data governance frameworks. As a data and analytics innovator, Stefanini is ahead of the curve in the ever-evolving ecosystem of data governance solutions. Contact us today!
