RingCentral Earnings July 23: All Three Things the Guide Did Not Pin Down Came In Ahead

---

Margins, Free Cash Flow and AI Traction. All Three Delivered.

UPDATE (August 7, 2026): RingCentral reported on July 23, and this page named exactly the right three variables. The guide was so tight that revenue and EPS were nearly pre-spoiled, so we said the action sat in margins, free cash flow and AI traction. All three moved the right way.

Margins. Non-GAAP operating margin came in at 23.4% (GAAP 7.7%), with full-year non-GAAP operating margin now guided to 23.6-24.0%.

Free cash flow. $180 million in the quarter, a 27.4% margin, up 25% year over year, with the full year guided to $615-625 million. For the value-trap debate this page described, cash generation of that shape against a heavy debt load is the single most important line on the page.

AI traction. ARR from paid AI products doubled year over year to 13% of total ARR, with net retention on those products above 100%.

The pre-spoiled lines beat anyway. Revenue of $657 million, up about 6%, cleared the $651 million expected, subscription revenue was $634 million, up 5.8%, and non-GAAP EPS of $1.22 beat the $1.17 quoted below. ARR reached roughly $2.8 billion, up 7%, with net monthly subscription dollar retention above 99%. Full-year guidance was raised to $2,635-2,646 million of revenue and $4.96-5.10 of non-GAAP EPS, and the company increased its dividend.

The debate is not settled, but it moved. Growth is still mid-single-digit and the debt is still there. What changed is that the AI bet this page called the turnaround thesis now has a number attached to it instead of a slide.

---

TL;DR

• RingCentral reports after Thursday's close, July 23, call at 5:00pm Eastern. The Street wants about $651 million in revenue and $1.17 in EPS, right against the company's tight $648 to $653 million guide.
• The guide is so narrow that the print itself is nearly pre-spoiled. The action is in three things the guide does not pin down: margins, free cash flow, and AI traction.
• RingCentral is the market's favorite value-trap debate: a cheap multiple and real cash flow, wrapped around a heavy debt load and decelerating growth.
• Whether the AI story flips it, and the trade.

The Board

When guidance is this tight, the revenue line is not the surprise. The surprise is everything underneath it.

Cheap For A Reason, Or Cheap For Now

RingCentral sells cloud phone and communications software (UCaaS), a category that grew fast, then matured, then got scary as the market wondered whether AI would eat seat-based software entirely. That fear, the same one that put enterprise software on trial after IBM's software-driven crash, is why RNG trades at a low multiple despite generating real free cash flow.

The bear case: growth is decelerating, a large debt load eats the cash flow, and the core seat business is exposed to exactly the AI-replaces-software worry hanging over the sector. The bull case: that fear is already in the price, the debt is being paid down with the very cash flow the multiple ignores, and the new AI products are about to show up in the numbers.

The AI Line Is The Whole Catalyst

RingCentral's turnaround pitch is that its AI layer (agentic assistants, conversation intelligence, its AI-native contact-center product) reaccelerates growth and proves it is an AI winner, not an AI victim. Thursday's tell is not the in-line revenue; it is any hard number on AI product adoption or contribution. Management naming a real AI revenue or bookings figure is the catalyst that re-rates a value trap into a turnaround. Vague enthusiasm without numbers leaves it cheap for a reason.

The Options Angle

• Options on RNG price a meaningful move because the stock is a low-multiple, high-debt name where sentiment can flip hard. With revenue near pre-known, the move keys off margin and AI commentary, not the top line.
• Trade free cash flow and net leverage, not the EPS penny. For an indebted software name, the deleveraging trajectory is the equity story. Debt down plus growth stable is the quiet buy.
• A reacceleration is a re-rate, not a beat. If you want the upside, you are betting the multiple expands on the AI narrative, which is a bigger prize and a bigger risk than a one-cent EPS beat. Structure with defined risk.

The One-Line Read

RingCentral is a cheap, indebted software name the market has left for dead, so the tight revenue guide is a distraction and the real print is the free cash flow and the first hard AI numbers, because those, not the penny of EPS, decide whether it is a value trap or a turnaround.