AudioCodes in the first quarter: ARR grew, but operating leverage has not arrived

AudioCodes opened 2026 with a quarter that strengthens the commercial growth story, but still does not resolve the quality-of-growth question. Revenue rose 2.9% to $62.1 million, and the Live and Conversational AI activities reached ARR, annual recurring revenue, of $80 million, up nearly 20% from the prior-year quarter. That is the working side of the story: Voice AI is growing quickly, Voca CIC posted record quarterly revenue, and the company continues to shift the center of gravity from communications gateways and endpoint equipment toward services, cloud, and intelligent voice applications. Still, profitability is not moving at the same pace: GAAP operating margin fell to 5.4% from 6.0%, and non-GAAP operating margin fell to 7.7% from 8.9%. Cash flow did answer one question left open at the end of 2025, with a sharp decline in receivables and $12.8 million of operating cash flow. The blocker remains on the other side of the cash account: CAPEX, dividends, and buybacks left a $7.4 million cash gap in the quarter, so the next quarters need to show not only more ARR, but also better services margins and an ability to fund capital returns without further liquidity erosion.

What the company is really selling now

The company is no longer just a supplier of enterprise voice equipment. It sits on the enterprise migration to Microsoft Teams, Webex, Zoom, cloud-based contact centers, and Voice AI applications, and sells around that base a mix of connectivity products, software, managed services, and professional services. That makes its economics both a growth machine and a margin machine: it needs to expand recurring revenue, but also prove that the new layer does not require ever more selling, implementation, and support labor.

The previous annual analysis framed the issue clearly: services were already more than half of revenue, but operating leverage had not arrived. The first quarter does not close that debate. It does show that legacy products are not disappearing immediately, and that the Live and Voice AI engines are gaining pace. But it also shows that the transition remains costly: operating expenses are growing faster than revenue, and services gross margin fell exactly when management is talking more about ARR, AI, and managed services.

That point matters because $80 million of ARR sounds like a strong software frame, but it still does not replace the income statement. The market will need to see that figure reach actual revenue, and that the revenue leaves more profit on the way to operating income.

ARR is growing faster than the reported business

The commercial headline for the quarter is better than the operating result. Management points to progress in Live managed services for UCaaS and CX, and to more than 50% growth in the Conversational AI business. Voca CIC, the Microsoft Teams-certified contact center solution, reached record quarterly revenue, while VAIC and Live Hub won new customers and expanded with existing customers. Meeting Insights continues to draw customer interest, including its on-prem version for organizations that want more control over data, service availability, and cost.

Inside the financial statements, however, that transition still looks much more modest. Product revenue rose 1.3% to $28.1 million, while services revenue rose 4.3% to $34.0 million. Services are growing faster, but not fast enough to change the profitability profile by themselves. More importantly, product gross margin jumped to 64.8% from 60.3%, while services gross margin fell to 67.3% from 68.6%. In other words, the improvement in total gross margin came mainly from products, not from the new engine the company wants the market to value.

The operating problem sits below gross margin. R&D expense rose 7.9%, sales and marketing rose 6.0%, and total operating expenses rose 6.4%, more than twice the pace of revenue growth. GAAP operating income therefore fell to $3.4 million, and non-GAAP operating income fell to $4.8 million. Net income dropped more sharply, to $2.0 million from $4.0 million, mainly because financial income moved from a $1.7 million gain to a $0.4 million expense. This is not an operating collapse, but it is also not evidence that the new growth layer is already creating leverage.

Cash flow improved in the right place, but returns still outran it

The good news in the quarter comes from working capital. At the end of 2025, one key question was whether the rise in receivables was timing or a sign that the new growth model needs more cash support. The first quarter delivered a good answer: net trade receivables fell by $9.7 million, and the company generated $12.8 million of operating cash flow, close to $13.5 million in the prior-year quarter. Deferred revenue also helped, with a $4.0 million cash-flow increase.

That strengthens the picture relative to the prior capital-return analysis. The quarter does not look like accounting profit that fails to turn into cash. Cash came in mainly from exactly where it needed to come in. Still, the all-in cash picture is less comfortable than the positive operating cash flow headline. The cash frame here starts with reported operating cash flow, which already includes the operating lease movement, and then tests what remains after CAPEX, dividends, and buybacks.

The company repurchased 1.74 million shares for $13.7 million during the quarter and paid a $5.3 million dividend. Together with $1.2 million of CAPEX, operating cash flow did not cover all of those uses. As a result, the broad liquidity measure used by the company, cash, deposits, marketable securities, and long-term financial investments, fell to $68.1 million from $75.7 million at the end of 2025.

This is not a liquidity-stress signal. The company does not have a heavy financial-debt story here, and the business still generates cash. But it is the same strategic choice from the prior cycle: using the balance sheet to return capital quickly before the services and AI engine has proven it can fund that pace on its own. As of March 31, 2026, only $1.6 million remained under the October 2025 Israeli court approval, which was valid through April 27, 2026. The point is not that the company cannot seek a new approval, but that the return pace in the quarter was rapid and not symbolic.

What will decide the next few quarters

2026 currently looks like a proof year, not a breakout year. The positive side is that the quarter provides three real signals: ARR is growing faster than revenue, Conversational AI is growing at a high rate, and receivables declined instead of continuing to pressure cash flow. These are not minor data points. They say the business transition is progressing and that part of the cash concern from late 2025 received an answer.

The unanswered part is earnings quality. As long as services grow with a lower gross margin, and sales and R&D expenses grow faster than revenue, it is hard to treat ARR as an engine that has already changed the company’s economics. The thesis will strengthen when services show both growth and a more stable margin, and when operating income returns to growth without leaning on a lower share count. It will weaken if the company keeps posting strong AI and Live headlines while remaining with a declining operating margin and capital returns that continue to consume more cash than post-CAPEX free cash flow.

The counter-thesis is fair: the current investments may be the right price for building a larger business, and the growth in ARR and Conversational AI may reach profitability with a lag of several quarters. The decline in receivables also supports the possibility that the business is healthier than the end-2025 balance sheet suggested. But the first quarter still does not provide full proof. Over the coming periods, the market is likely to focus less on AI terminology itself and more on the connection between ARR, services margins, operating expenses, and free cash flow after capital returns.