For the past few years, the UC and CX sectors have been in a state of suspended animation, waiting for AI to graduate from a mesmerizing demo to a business-critical utility. If the latest 8×8 earnings results are a reliable barometer, that moment of graduation has arrived, at least to an extent.
While the company reported a solid return to top-line growth with $185 million in total revenue, the headline figure disguises a far more radical shift in the plumbing of the enterprise tech stack. We are possibly witnessing a fundamental bifurcation in how companies consume technology, moving decisively away from the static safety of subscriptions toward dynamic, usage-based models driven by the operationalization of AI.
The most telling metric from the earnings call was the composition of the revenue, rather than the total. Usage-based offerings, primarily CPaaS APIs and AI solutions, surged nearly 60 percent year over year and now account for more than 20 percent of service revenue. In an uncertain economic climate, the appetite for locking into massive, multi-year seat licenses for unproven technology has evaporated. Usage-based pricing aligns the vendor’s revenue directly with the customer’s success; if the tool works, usage spikes, and the vendor gets paid.
Samuel Wilson, 8×8 CEO, framed this culture change explicitly on the call:
“The pay-as-you-go approach appeals to customers because it reduces risk as they adopt new technologies. It also raises the bar for vendors. Revenue is linked directly to successful customer outcomes and business activity instead of long-term subscriptions that may or may not be implemented. We believe this is the way of the future.”
Perhaps most significant is the specific nature of this usage. 8×8 reported that Voice AI interactions increased by more than 200 percent, now representing the vast majority of all AI interactions on their platform.
This is a crucial distinction for the skeptical CTO. Text-based chatbots are relatively low-stakes and easy to deploy; Voice AI requires low latency, high accuracy, and deep integration. A 200 percent spike suggests that the technology has finally crossed the reliability threshold required for complex production environments, such as healthcare scheduling and biometric authentication in finance, rather than just handling password resets.
The Earnings Dividend of Discipline: Breaking Down 8×8’s Financial Health
Turning to the financial health of 8c8 itself, the Q3 report paints a picture of a company that has successfully navigated a challenging but necessary period of restructuring. 8×8 delivered its twentieth consecutive quarter of positive operating cash flow, a streak that underscores operational resilience in a high-interest-rate environment.
Since August 2022, the company has aggressively deleveraged, reducing its debt principal by $224 million. For the enterprise buyer assessing vendor viability, this balance sheet discipline is a reassuring signal that the partner will be around to support the long-term roadmap.
The quarter also marked a significant operational milestone: the completion of the Fuze customer upgrades. The legacy Fuze platform, acquired years ago, is now officially decommissioned. While the forced migration created a temporary revenue headwind due to inevitable churn, the strategic upside is significant. 8×8 is no longer splitting its R&D budget between maintaining a legacy stack and building the future. Every dollar of innovation spend now flows into the single platform that customers are actually using. This elimination of technical debt is the quiet engine of future margin expansion and feature velocity.
The Hybrid Paradox Behind 8×8’s Earnings
Arguably, the broader implication for the market is a challenge to the prevailing narrative around automation and the workforce. The industry has spent years debating the “Death of the Seat” with a tone bordering on the apocalyptic.
However, 8×8’s data suggests a counterintuitive reality. 8×8’s AI traffic is up 200 percent, while its human contact center seat count is also growing quarter over quarter. We are likely looking at a real-world manifestation of the Jevons Paradox: as AI potentially makes customer interactions cheaper and more efficient, the demand for interactions increases, possibly creating more work, not less.
Again, this is just one signal, and we can’t get ahead of ourselves, but it is tangible and substantial. It suggests the future of the contact center could potentially be not one of replacement, but of elevation. If an AI can handle the micro-transactions, such as scheduling, payments, and verifications, human agents are left with the tasks that require empathy, complex problem-solving, and more meaningful revenue generation.
