According to Intercom CEO Eoghan McCabe, Intercom is a product-first company, which means it always leads with its product. But at its Pioneer summit last week, McCabe showed what the vendor is working on now to get companies prepared for what’s coming. Fin will become a single, unified AI agent capable of handling the entire customer experience from start to end.
The Customer Agent
McCabe shared two key ideas. First, he says multiple agents will destroy the customer experience. The complexity of competing agents with varying priorities, goals, contexts, communication styles, and channels will pose serious challenges for organizations. Second, a truly exceptional customer experience is finally possible. Today’s agents can handle many different use cases across the customer journey:
Fin will become the thing that makes contact with the customer for the very first time; holds their hand through consideration, and purchase; be there with them at every step; be their expert personal concierge and know everything about their life with your business; to help businesses achieve the goals that most matter to you. All at a level of speed and attention and detail we’ve never seen before.
What does this mean exactly? Fin is going to get a ton of new capabilities:
Multiple roles that can fluidly and dynamically move between and blend together. Each role will be a world-class expert.
Goals to pursue. Tell Fin your objectives and priorities (e.g., customers, company, revenue) and Fin will pursue them. Of course, safety rules and supervision will be included.
Memory that persists and grows over the entire customer lifecycle. Fin will have deep customer context.
Knowledge of the business, its products, policies, processes, history, and plans.
Interoperability with different tools, systems, and channels. It will also be able to collaborate with different AI agents.
Yes, the bulk of the company’s resources will continue to be in customer service for a while, but McCabe said the new Customer Agent experience is critical to get right.
How did Intercom get here?
Intercom Co-Founder and Chief Strategy Officer (CSO), Des Traynor, shared some insights into the company’s new vision in his session, “Convergence and Transformation in the Post-AI World.” Traynor argues that AI is a convergent force, breaking down boundaries. It can take things that were once separate and bundle them together to build hyper-specific tools and products. This has led to a ton of products, which he refers to as an ‘Appsplosion’:
Things were separate for a reason. They had different audiences and owners. The user interface was different, the backend was different. There are many reasons. But these differences are starting to fade away.
Traynor suggests convergence is coming to both the user interface and the code layer. He believes we’re moving from searching to doing, and customers will test the boundaries of your product. Customers will expect your product to do more things. Then there’s the AI CoPilot. It’s a nice sidebar to help you do things quicker in a product, but it’s eroding the need to be good at software. There will also be a lot of integration that needs to happen, but won’t be as hard as it used to be, partly because code is becoming easier to generate using AI.
We are going to start seeing the landscape change, Traynor explains, with software merging or disappearing, especially smaller tools. If and when products converge, functions will converge too. Today, every function is defined by roles, goals, and product. But roles aren’t static; they also evolve, either disappearing because they are no longer needed, dispersing because everyone can do them, or merging where one person will do multiple jobs. We’ll start to question why things are separate and look at ways to bring them together.
The question Traynor poses is what will happen to Customer Support, Service, and Sales roles? He said there are many different roles within the industry (e.g., customer support, sales development representative, customer success, customer onboarding) and there are many similarities between them. For example, they have the same product knowledge, policies, and tone of voice. The similarities are core to the business. The differences are local, and this is where AI comes into play. AI speaks every language, works in every timezone, has the same salary, skillset, and knowledge.
Traynor says we will start to see convergence at the local scope, where products expand to adjacent functions. He cites Notion and Grammarly as examples of this happening today. We’re seeing this convergence happen across organizations. In development and engineering, AI will be fully authorized to work across every part of the codebase, and engineers will become fluent in it as well. Likewise, in marketing, we’re seeing AI enable the rise of the full-stack marketer.
Then there’s customer-facing experience. Traynor asks if we need a different agent for every interaction or if it makes more sense to have an agent who knows everything.
There is an opportunity in front of customer service organizations and their people, he said. You’re not just good at AI for customer service. It’s more likely that you are good at AI, transforming businesses, transforming customer conversations, and delivering insane results. And that ability reaches across the organization and the customer lifecycle, not only in customer service. Traynor said that AI lets people go deep in all areas, and the people who will benefit the most are the ones who adapt the quickest.
My take
Traynor makes a lot of great points in his talk. Roles that used to be function or department-focused are shifting, and there are opportunities to become more of a generalist with the help of AI. This is happening in Marketing, Sales, and Customer Service.
And that’s not really new. We’ve talked about roles converging before to enable a better customer experience. Technology has been a challenge in achieving that, as have company politics. AI is now making that shift more of a reality, and the vision of a customer AI agent that can support a customer from start to end is a very good idea. It may not be unique, but it is a good idea.
If you look at what other tech vendors are doing, you see and hear similar things. Whether it’s HubSpot, Salesforce, or another tech provider, the goal is to understand the customer at a deeper level than during the current visit or call, and have that context when the AI engages with them.
That means knowing everything about them and their relationship with the company (and possibly other external data). Does that mean only one agent? Maybe. It may mean one customer-facing agent, which then interoperates with other agents on the backend. Or it may mean multiple customer-facing agents who have the same data and context of the customer, ensuring a consistent experience across agents.
The end goal is great Customer Experience that helps customers solve their problems, answer their questions, and do their work. AI can help us do that. How that agent experience is built will be different for different vendors. What we know now is that Intercom is focusing on a single agent, and it will be interesting to see how that agent experience evolves as they build it.