AI Demand Is Firming; Consumer Demand Still Has to Prove It

The market’s cleanest tell is technology—not the whole basket

The hypothesis that earnings growth and resilient demand can support DDOG, SNOW, CRWV, AN, META, RH, WSM and ETH is getting a clear but divided test. Wednesday’s tape favored technology: QQQ gained 2.31% and software ETF IGV rose 1.93%, against a 1.32% gain in SPY and 0.52% in DIA at the 16:00 ET close. CRWV surged 24.71% on the day, while META fell 1.97%. Those are closing prints from FMP, delayed 15 minutes; they are a snapshot, not proof that the same demand conditions apply across every business.^[1]

Reuters tied Wednesday’s technology-led advance to CoreWeave’s results and renewed confidence in AI spending. That explains the day’s leadership better than a broad claim about consumer resilience.^[2]

CoreWeave: strong demand, with a financing and execution bill

CoreWeave’s third-quarter revenue was $3.661 billion, up from $1.365 billion a year earlier. The company reported $132.5 billion of revenue backlog and lifted its 2026 revenue outlook to $13.0–$13.3 billion. It also raised its capital-expenditure range to $37–$39 billion and reported $846 million of quarterly net interest expense. Net loss was $709 million, despite positive operating income. Those figures support the demand case—but also show why backlog, deployment timing, cash needs and financing costs belong in the same sentence as growth.^[3]

The first market reaction was emphatic: CRWV closed up 24.7% on November 11, then traded at $61.44 in after-hours trading at 16:07 ET, 1.10% above the close. The live print is not a second full-session return; it is a modest incremental move after the close.^[1]

Software adds a broader-demand check

Datadog’s Q3 revenue rose 35% year over year to $1.197 billion. Management said growth excluding AI customers reached the high 20s and accelerated again excluding its largest customer, whose usage declined but remained above its minimum commitment. It also counted more than 850 AI customers, up from about 750 in the prior quarter. This is meaningful evidence that the usage story extends beyond one large AI account, while that customer’s lower consumption remains a real concentration caveat.^[4]

Snowflake’s latest available call in this research pass was fiscal Q2 2027. Product revenue grew 37% year over year for a third straight quarter of acceleration, with management citing AI workloads and healthy expansion. But the same update lowered the expected product gross margin to 74%, attributing the change to a higher mix of AI workloads that currently carry lower contribution margins. Growth is arriving; the question is how efficiently it converts into profit.^[5]

Consumer and home-furnishings evidence is more mixed

The names outside software and AI infrastructure do not yet provide the same uniformly strong confirmation. AutoNation reported Q3 adjusted EPS of $5.11, while a news report described the result as slightly below the $5.17 analyst forecast; the company highlighted record after-sales gross profit and 6% customer-pay growth. That is a mixed operating picture, not a pure demand read-through.^[6]

For RH, the latest evidence found here is its Q2 report and outlook: sales rose 2.6% year over year, but the next-quarter revenue outlook was reported below analyst expectations. Williams-Sonoma’s latest available earnings evidence is stronger: Q2 comparable brand revenue grew 6.2%, with a 17.3% non-GAAP operating margin. These are different quarters and business models, so they should not be collapsed into a single consumer-demand conclusion.^[7]

Ethan Allen’s first quarter ended September 30 showed a sharper caution signal. Sales declined 11.8% year over year; adjusted operating margin fell to 2.2% from 7.2%; and management cited lower design-center traffic, a soft housing market and tariffs. Retail written orders improved sequentially and were up 0.3% in October, a potentially stabilizing detail—but one month does not erase the weaker quarterly comparison. The prompt names ETH; the company’s official release identifies Ethan Allen as NYSE: ETD, so the symbol mapping should be verified rather than assumed.^[8]

![A carefully furnished living room with a neutral-toned interior.]

Macro sets a less forgiving hurdle

The latest macro snapshot available here, for October, combines 4.1% unemployment and 2.07% year-over-year real GDP growth with 3.72% CPI inflation, a 4.88% 10-year Treasury yield and a 19.92 VIX reading. Consumer sentiment registered 48.1, down 12.7% year over year. This is not a recession reading in the tool’s classification, but it is a mixed backdrop: growth has not vanished, while elevated yields and weak sentiment make it harder to assume discretionary spending will accelerate broadly.^[9]

What to watch next

• Conversion, not just commitments, at CRWV: track whether contracted backlog turns into delivered capacity and revenue while capex, financing costs and operating margins remain manageable.
• Breadth and concentration in software: DDOG’s next updates should show whether non-AI customer growth stays healthy and the large-account usage reduction stabilizes; SNOW’s next report can test whether AI-led consumption growth improves margins over time.
• Fresh consumer evidence: compare new AutoNation commentary with upcoming RH and Williams-Sonoma results rather than extrapolating from older-quarter comps. For home furnishings, follow traffic, written orders, housing conditions and tariff effects.
• The macro hurdle: yields, inflation and sentiment matter most for the rate-sensitive and discretionary portions of this basket.

The balanced read is that earnings and demand evidence support parts of the one-year thesis, particularly AI infrastructure and cloud software. The evidence does not yet support treating all eight names as beneficiaries of one broad, resilient-demand cycle. That would require continued software usage growth, successful infrastructure delivery and a clearer stabilization in consumer and home-furnishings demand. This is research, not trading advice.