The Phone Still Rings. The Revenue Still Leaves.

Every service business owner I talk to knows they're losing jobs to missed calls. What very few of them realize is how systematically it's happening, or how fast the gap between "has AI-assisted follow-up" and "doesn't" is about to widen.

This week, Five9 agreed to pay $325 million for Level AI, an AI-native platform that handles more than one billion customer interactions a year for companies including Gusto, Chime, and VistaPrint. That deal landed within days of Salesforce completing its Cresta acquisition and a Dialpad/Ada merger, meaning three of the largest contact-center infrastructure plays of 2026 all closed inside a two-week window. The overlap isn't a coincidence. Enterprise software vendors are racing to own the AI layer around customer conversations, and they are paying for it at acquisition prices that reflect how much revenue they believe is sitting in that layer.

Here is the question service-business operators should be asking: if billion-dollar companies are spending hundreds of millions to fix the customer-conversation problem for the enterprise, what does that tell you about how much the problem costs at the smaller scale you operate at?

The answer isn't comfortable. A solo operator or a business with two or three people covering phones loses more than a lead when a call goes to voicemail at 6:47 PM on a Thursday. They lose the job, the review, the referral, and the relationship, all in one unanswered ring. Research from the Harvard Business Review found that companies responding to inquiries within an hour are seven times more likely to qualify the lead than those who wait even a single hour longer. For a service business booking $250 to $2,000 per job, that number deserves serious attention.

The Five9/Level AI deal is specifically interesting because Level AI built its own proprietary model family, called Latitude, trained specifically on customer conversations rather than general text. That matters because a model trained on how real service interactions unfold, including objections, pricing conversations, scheduling, and complaints, produces meaningfully different outputs than a general-purpose AI wrapper. The fact that Five9 paid a reported four times Level AI's total funding history for that capability is a strong signal that purpose-built conversation AI is not a commodity.

None of which means a service-business owner needs an enterprise contact-center platform. What it means is that the same fundamental problem has clear solutions at every price point, and the gap between operators who have automated their first-response layer and those who haven't is now measurable in booked revenue.

A practical starting point: map every touchpoint where an inquiry can arrive (phone, text, web form, Google Business Profile, Instagram DM, email) and ask which ones get a response within five minutes if they arrive at 7 PM on a weeknight or at noon on a Saturday. The ones that don't are your revenue leak. The CX AI tools now available to small operators, including AI-assisted SMS follow-up, voice AI for after-hours calls, and automated booking flows, don't require enterprise pricing or a 90-day implementation. Most are operational within a week.

The M&A wave confirms something important. The largest companies in customer-experience software have concluded that AI-assisted conversation handling is not a feature they can afford to leave to a third-party add-on. They're buying it outright because the revenue it protects is too large to leave at risk. That same logic applies at your scale. The revenue at stake is different in dollar amount, but the percentage of bookings at risk is often higher, not lower, for a business without a 24-hour team covering their phones.

The consolidation isn't the story. The revenue leaking out of your business right now, one unanswered call at a time, is the story.