Personal AI agents are not waiting for businesses to figure out what to do with them. Meta’s Muse has reached 10 million US downloads just 40 days after launch. Muse can browse, fill out forms, negotiate, and keep working toward outcomes on a user’s behalf. It’s the splashiest example yet of a growing class of AI agent that can act on consumers’ behalf.   

So, what happens when your customer sends an AI agent to contact you instead of doing that themselves? 

For decades, brands have been able to externalize some of the cost of service onto customers in the form of effort. 

Consider what happens today when a customer doesn’t get what they need when they contact you. Maybe the chatbot sends them in circles. Maybe they call, wait 20 minutes, explain the problem, and still don’t get a resolution. Sometimes they try again. Sometimes they just give up.  

Making another call costs the customer time and effort. So does escalating a dispute, chasing down a refund, or fighting a cancellation policy. Eventually, getting the issue resolved may simply stop being worth the trouble (that is, the effort).  

What happens when persistence gets cheap? 

Personal agents give customers a way to hand back that “effort” bill. Real talk: Contact centers have never seen the full volume of customers’ unresolved needs. They have only ever seen the portion of customers who are willing to spend their time pursuing what they wanted to ask or get resolved. If my personal agent can chase a $40 billing error while I go about my day, I don’t have to decide if $40 is worth an hour of my time.  

Contact centers have spent decades investing in containment and automation to manage demand. When those strategies genuinely resolve the customer’s need, they work beautifully. But when personal agents expose unresolved demand that customer effort used to keep out of sight, as a contact center or CX leader, don’t just look for new ways to contain it. Use it to find the problems customers shouldn’t need to chase you about in the first place. 

Find the demand customer effort is hiding 

Personal agents will increase service volume. Vendors will soon offer plenty of technology to identify, route, or even block personal agent traffic. But knowing that an AI agent made the contact tells you very little about what created the demand. Before you build a strategy around this new source of traffic, pressure-test the demand you may actually have to serve: 

  • Find where customers give up. Many brands include channel abandonment (e.g., hang-ups) in their containment measurement. Separate out channel abandonment and examine which intents routinely end without evidence that the customer got what they needed. Demand that disappears when customers give up today will come back when AI is doing the pursuing. 
  • Find where customers have to persist. Identify the issues that require repeat contacts, escalations, channel switching, or multiple attempts before customers finally get an outcome. The more effort that resolution requires today, the more demand personal agents could create by making that persistence cheap (i.e., effortless) for that consumer. 
  • Revisit issues whose low volume may be misleading. Contact volume alone can lead you to prioritize the wrong improvements. Compare known recurring issues with the total number of customers affected, not just those who contact you. A problem that affects many customers but motivates few to contact you may not stay low-volume when customers can delegate the effort of pursuing it. 
  • True-up resolution measurement. Don’t assume a closed case or the absence of a repeat contact means the customer got what they needed. Use conversation intelligence alongside downstream outcome data to validate resolution. Personal agents will expose the gap between what you define as “done” and whether the customer agrees. 

Want to understand how personal agents could reshape service demand in your organization? Book a guidance session with me!  

Thank you to Max Ball and Kate Leggett for contributing to this blog. 

Share