Your Helpdesk Already Talks to Machines — It Just Doesn't Know It Yet
Call Centre Helper published an interview on 31 August 2026 with Matt Lyles, a keynote speaker, brand consultant and podcast host, on what the site calls "machine customers" — software, AI agents or connected devices that act on a human's behalf when dealing with a business. Lyles traces the pattern back to Amazon's Dash buttons, launched in 2015, which let shoppers reorder a single product with one press before being superseded by voice ordering through Alexa. His argument for contact centres: this isn't a future problem, it's a labelling problem — you're probably already serving machine customers and treating them like confused humans.
What exactly counts as a "machine customer"?
Any technology acting on behalf of a person — from a connected printer reordering toner to an AI agent researching a purchase — not an autonomous robot buying independently.
That distinction matters for support teams because it changes what "resolution" looks like. A printer pinging a manufacturer about low ink doesn't want empathy, hold music or a satisfaction survey — it wants a clean API response. If your support stack routes every inbound contact through the same chat widget and the same CSAT prompt regardless of who, or what, initiated it, you're measuring the wrong thing for an increasing slice of your volume.
Is this actually happening in queues today, or is it still hype?
Partly hype: Lyles is clear that autonomous machine purchasing isn't here yet — AI mostly does research and recommends, humans still approve.
He put it directly: "We're not yet at a point where we have autonomous robots buying on our behalf." The practical version showing up now is AI doing the comparison-shopping homework and humans clicking confirm. For support teams, that means the contact you're getting might be a human who's been primed by an AI summary before they ever open a ticket — arriving with sharper, more specific questions than your macros were written for, even though no machine actually reached your system.
How much of this is genuinely new versus repackaged?
Little of it is new — replenishment automation dates to at least 2015; what's changed is the sophistication of the research layer in front of it.
That's worth remembering given how this site has flagged vendor language before — most recently in a piece on why, "when every vendor sounds excellent, polish stops being evidence." "Machine customer" readiness is exactly the kind of framing that can be sold as a brand-new capability when the underlying mechanic — automated reordering, rules-based triggers — has been sitting in loyalty and subscription platforms for a decade.
What should you actually ask your support vendor?
Ask how their platform distinguishes a machine-originated contact from a human one, and what happens differently once it does.
Push for specifics: can the platform detect API-originated or bot-originated contacts separately from chat sessions, does it route them away from CSAT surveys designed for people, and does anyone measure resolution time for those contacts differently? If the answer is "it all goes through the same queue," that's not a machine-customer strategy, it's an unexamined assumption — the same gap this site raised in its Avaya coverage on CX vendors splitting over AI roadmaps.
Frequently asked questions
Do I need new software to handle machine customers?
Not necessarily new software — but you likely need new rules inside what you already run, separating automated and human-approved contacts from routing logic built for people.
Is this the same as a chatbot handling tickets?
No. A chatbot answers a human. A machine customer is the thing initiating contact on a human's behalf, which is a different routing and measurement problem.
Where does this connect to what this site has already covered?
It sits alongside recent pieces here on CCaaS evaluation and vendor claims — the consistent theme is that new AI framing needs the same scrutiny as any other vendor metric.
Source: Call Centre Helper, "Are You Ready for AI Customers?", 31 August 2026.
Learn what machine customers mean for contact centre customer experience and how to prepare your organization for the next wave of Artificial Intelligence.