Cisco’s fiscal 2026 results answered the easiest question about its AI exposure: customers are placing real orders and the company is converting part of that demand into revenue. The harder question is whether the growth can persist long enough—and at sufficient margins—to justify a valuation near the top of Cisco’s recent history. That makes the next earnings report less about proving that AI demand exists and more about the quality and durability of the expansion.
Orders have moved beyond a single headline
Cisco reported fiscal fourth-quarter revenue of $17.3 billion, up 18% from a year earlier, with networking revenue up 28%. Product orders rose 35%; excluding hyperscalers, they still increased 25%. Networking orders grew 40%. Those figures matter because they show strength across more than the handful of cloud companies buying the largest AI clusters.
For the full year, Cisco generated $63.3 billion of revenue, up 12%, and $14.2 billion of operating cash flow, roughly flat. Networking contributed $34.7 billion, compared with $28.3 billion a year earlier—an increase of $6.4 billion, or about 22.6%. Services stayed around $15.0 billion. The mix therefore shifted toward product revenue, explaining why investors must watch gross margin even while headline growth accelerates.
Cisco says fiscal 2026 AI-infrastructure revenue reached $4 billion on $9.3 billion of orders, and it expects revenue from that category to reach $7.5 billion in fiscal 2027. If achieved, that would be an $3.5 billion increase, or 87.5%. It would also equal about 10.3% of the midpoint of Cisco’s $72.2 billion to $73.4 billion full-year revenue guidance. AI infrastructure is no longer immaterial, but most of Cisco’s revenue would still come from traditional networking, security, collaboration, observability and services.
Growth is not yet translating proportionally into cash
Fiscal 2026 operating cash flow was flat despite double-digit revenue growth. Some divergence can result from working capital, acquisition integration, deferred revenue and customer-payment timing, but the gap is a useful counterweight to the order numbers. Hardware growth can require inventory and supplier commitments before cash is collected. It can also carry lower gross margins than software or support.
Cisco’s non-GAAP operating margin was 34.8% for the fiscal year, demonstrating substantial profitability. Yet management warned of a slight gross-margin headwind as high volumes of hardware ship. If AI systems increase revenue but dilute gross margin or consume working capital, their contribution to free cash flow will be smaller than their sales contribution. Investors paying a premium multiple need evidence that software, optics, security and recurring support revenue expand alongside the boxes.
An order is also not the same as recognized revenue. Delivery schedules, customer acceptance and component availability determine when a booking becomes sales, while cancellation terms affect its quality. Cisco’s $9.3 billion of AI orders provides useful visibility, but the $4 billion of fiscal 2026 revenue shows that conversion unfolds over time. The spread can support future growth; it can also widen if deployment bottlenecks delay customer projects.
Guidance is strong—and already visible
For fiscal first-quarter 2027, Cisco guided to $18.0 billion to $18.2 billion of revenue and non-GAAP earnings of $1.32 to $1.34 per share. The revenue midpoint is 4.6% above the $17.3 billion reported in the preceding quarter, though sequential comparison is affected by seasonality. For the full fiscal year, the company expects $72.2 billion to $73.4 billion in revenue and non-GAAP EPS of $5.05 to $5.11.
At the midpoint, full-year revenue would be $72.8 billion, about 15.0% above fiscal 2026. Management also said comparisons become harder from the second quarter onward. That is the core risk in buying after a strong run: first-quarter performance can meet guidance while the market focuses on whether order growth is normalizing against a tougher base.
Splunk changes the quality of the comparison
Cisco’s reported growth includes the benefit of Splunk, so investors should separate acquisition-driven scale from organic demand. Splunk can strengthen recurring software, security and observability revenue and create cross-selling opportunities with Cisco’s network installed base. Integration can also create duplicate costs, sales disruption or slower-than-expected synergy capture.
The most valuable disclosure is not another broad statement about cross-selling. It is a bridge showing organic product growth, remaining performance obligations, recurring revenue, segment margins and cash conversion. That would reveal whether Cisco is becoming a more valuable software-and-services compounder or simply benefiting from a hardware cycle and a larger consolidated revenue base.
What the current valuation requires
Trefis estimated on October 7 that Cisco traded near 7.4 times annual sales and 35.1 times earnings, close to the upper end of its ten-year sales-multiple range. Those are market estimates, not company guidance, and they can move quickly. Their importance lies in the expectation they represent. A high multiple can be sustained if AI revenue grows toward $7.5 billion, broader orders remain healthy, margins hold and cash flow catches up. It becomes difficult to defend if growth slows just as hardware mix weighs on profitability.
The bull case is not merely “AI.” It is that Cisco uses an unusually strong networking cycle to sell more silicon, optics, security and observability into a larger installed base, while Splunk increases recurring revenue. The bear case is that customers over-order infrastructure, growth decelerates against difficult comparisons and the premium multiple compresses before cash generation improves.
A sensitivity illustrates the operating leverage investors are underwriting. At the $72.8 billion revenue midpoint, each one percentage point of operating margin represents about $728 million of annual operating income before tax and other items. That does not predict a margin change, but it shows why mix matters. A one-point margin shortfall could offset a meaningful portion of incremental AI revenue, while a one-point improvement would make the same sales growth considerably more valuable.
The $3.5 billion implied increase in AI-infrastructure revenue deserves the same discipline. If it arrived at a margin below the company average, consolidated sales could beat while profit quality disappoints. If customers also adopt security, observability and support products, the lifetime economics would be stronger than the initial hardware margin suggests. Segment attach rates and renewal behavior are therefore essential to judging whether AI demand is creating a durable franchise or a cyclical equipment surge.
Balance-sheet deployment adds another variable. Strong profitability gives Cisco room to invest, repurchase shares and integrate acquisitions, but every dollar allocated to inventory or another deal competes with shareholder returns. A cleaner conversion of earnings into cash would reduce that trade-off and support the premium investors are currently assigning to growth.
The next report’s decision points
Three outcomes would materially strengthen the thesis: revenue above the $18.2 billion top of guidance, product orders that remain robust excluding hyperscalers, and gross-margin performance that absorbs the hardware surge. Cash flow and remaining performance obligations would add important confirmation.
A result inside guidance would not be poor. It would simply confirm what management and analysts already expect. After the stock’s large advance, Cisco needs more than a respectable quarter; it needs a credible path from orders to profitable recurring revenue. The company has produced the demand evidence. Fiscal 2027 must now show that the economics deserve the valuation.