Get Ahead and Equip Your Team for AI Agency in Customer Service

Summary

Google, OpenAI and others have stated that they will imminently release bots that have agency — acting on behalf of humans to book flights, manage subscriptions, and gather information. Unlike humans, AI can juggle many interactions at the same time without fatigue, leading to a huge surge in automated queries for businesses. These bots will have a multitude of webchats and calls with brands simultaneously without any concern for the length of engagements. Impacts on revenue and customer service will be significant for businesses that are not prepared.

Actionary Take: Brands that are not prepared for bots with agency will struggle to provide effective customer service, and will face new sales and marketing challenges.

Enhancements in AI Are on the Cusp of Converting Chatbots Into Autonomous Agents

In 2018 Google Duplex was revealed to the world. Google’s vision was to launch an intent-based AI solution to accomplish real-world tasks over the phone [1]. It was unable to realize the ambition. CEOs of leading generative AI companies are now revisiting this ambition, targeting beyond human-like AI capabilities. Sam Altman, the CEO of OpenAI, recently made headlines stating that he wants AI to become a “super-competent colleague”. Dario Amodei, the CEO of Anthropic, in a recent interview with the New York times spoke of leveraging generative AI to make reservations, book an Uber, go to websites and talk to other people [2]. Gartner have written many research notes and even a book on the topic. Bots with agency are about to become part of our everyday lives and it will have a profound effect on how consumers engage with marketing, sales, and support. It’s critical to be ready for the change this creates. 

• Leading AI companies are embedding their technology into everyday consumer touchpoints. OpenAI’s partnership with Apple, Microsoft’s ongoing development of its Copilot+, Samsung’s Galaxy AI and others are early examples. Not only are these solutions acquiring agency, they are rapidly taking ownership of the digital touchpoints between brands and consumers [3][4][5].
• Large language models (LLMs) being developed by these companies will gain agency. Specifically, the ability to act autonomously across a more diverse range of tasks, in varying formats, in the next 12 months. When these solutions gain agency, they will act on behalf of the consumer, interacting with the brand.
• Customer Service leaders are focused on other automation abilities for their teams. AI Innovation providing customer self-service solutions and augmented staff experiences are the current focus. However, ChatGPT has already surpassed 180 million active users on their website. AI is being embedded across customer devices. Adoption is already sufficient to ensure that when AI gains agency the effect will be immediate and significant.
• Sales and marketing will have to adapt to service a bot-empowered consumer base. Agency will not be limited to customer service activities. Consumers are increasingly moving from traditional search to instead asking their chosen bot to find the answers for them. Consumer behaviors have already changed. 

Bots Acting on Behalf of Consumers Will Impact Customer Service in Significant Ways  

Bots provide customers with limitless time for minimal effort. Just as web search provided customers with the ability to reach a vastly wider range of brands; bots provide customers with limitless time, endless patience and a deep understanding of their rights and legal options. Customers will be able to gather information from more sources, to relentlessly chase a resolution if they’re unhappy, and to be far more selective in getting exactly what they’re after. This transformation will impact customer service teams in many ways.     

• Interaction volumes will increase enormously. Customer service organizations rely on consumers having limited time, and a reluctance to deal with “hassle”. AI is already present within most consumer touchpoints, has limitless time, and endless patience. It can contact multiple companies simultaneously to gather information, ask for discounts, or check stock. It can also chase relentlessly for an answer, and if that answer isn’t satisfactory, escalate internally, to an ombudsman or into small claims court. Having your own AI may deflect some of this, but effective implementation is critical. It doesn’t even matter if the consumer has never heard of your brand.
• The skills that matter will change. Instead of interacting entirely with humans, brands will have to interact with AI. AI is never emotional. The skillset required (from a human agent or an AI agent) to handle a human customer is different to that required to handle a human customer’s bot. A bot will be tasked by a customer to achieve a goal, its entire focus is on achieving that goal in the best possible way. If your organization isn’t setup to frictionlessly deliver that outcome (ideally without needing to involve a human agent at all), the bot will drag you through that friction with it – If the exact answer or resolution it’s after isn’t available online, or through your own bot, it’ll chase your live agents until it gets it.
• Interaction timing and response times will change. In a human consumer to brand interaction, the interaction takes place when the consumer realizes that they need to interact. Customer service organizations are setup to handle this, and are building in automation to handle this. Consumers have to adapt around this. However in a consumer’s bot to brand interaction, the consumer tells the bot to solve the problem, and leaves it to complete the task. The bot doesn’t need to sleep, or go to work, it doesn’t mind sitting on hold for an hour or calling back the next morning when you’re open. The content, timing, volume and nature of interactions will all change.
• Brands will have to learn to sell and market to bots. When consumers switch from searching for items themselves and instead ask their bots for what they want, the way brands are identified will change. SEO matters far less if the bot can search 500 pages of google results for the answer. SEO isn’t even a consideration if a bot uses its knowledge base, and not a search engine for an answer. An AI vendor can identify how satisfied consumers are with a product much more accurately than any review site. Beyond this, knowing who’s calling becomes critical. A bot reaching out has a completely different interaction profile to a human reaching out, and demands a different response.

Being Prepared for this Unprecedented Shift in Consumer to Brand Engagement is Critical 

Businesses should treat the impact of AI like any other risk to their business. They should analyze it, identify the potential impact it may have, and ensure they are prepared to handle that impact. Different industries, business models and geographies face significantly different overall risks. A North American retail organization providing phone-based support only, with a limited online presence and a chain of stores with different stock and pricing levels would face a very different risk profile to a boutique hotel with an online booking tool, a broad selection of customer reviews, a detailed website, and an extensive publicly available Q&A section.  

• Evaluate your specific level of risk. Assess the impact of unlimited customer interactions on your brand. Identify goals that are currently too effort-intensive for customers and evaluate the effort required if customers pursued these goals without constraints. Understanding this impact will help in developing strategies to manage increased interaction volumes and demands.
• Gather data specifically for bot consumption. Understand what problems customers are looking to solve, what topics they care about, and where they currently experience friction in readiness for less human-in-the-loop response. Ensure bots are provided with the information they need to achieve their goals. Enabling automated gathering of information will protect against spikes in demand. Ensuring you provide satisfactory outcomes to customers will reduce the pain from customers who aren’t satisfied with what they’re offered.
• Ensure your workforce are skilled in using AI. AI isn’t coming for your job, someone who knows how to leverage AI is. Train employees to leverage AI effectively. Employees who understand the capabilities and limitations of AI tools can better serve and sell to customers through these tools, ensuring the business remains competitive and successful in an AI-driven world.

References 

1. Google Duplex announcement: An AI System for Accomplishing Real-World Tasks Over the Phone
2. CEO of Anthropic interview: Ezra Klein Interviews Dario Amodei
3. OpenAI: OpenAI and Apple announce partnership to integrate ChatGPT into Apple experiences
4. Microsoft Blogs: Introducing Copilot+ PCs
5. Galaxy AI Launch Feature: Galaxy AI

Related Products and Services:

To speak to an analyst about this research, Book an Inquiry or Connect Now.