Cloud demand is holding up better than the housing-sensitive consumer

The opening tape is not validating a broad “resilient demand” trade. It is separating businesses with measurable software usage from businesses still waiting for housing turnover and confidence to improve.

The lead: a two-speed test of resilience

At 12:07 p.m. ET on September 28, the broad market was under pressure: SPY was down 0.77%, QQQ 1.22%, and DIA 0.67% from the prior close. DDOG was modestly higher, while SNOW, RH, WSM, LZB and LESL were lower. The moves do not prove a durable regime change, but they offer a useful opening read: investors are rewarding evidence of operating demand more readily than a general consumer-recovery narrative.^[1]

The hypothesis for this desk was that earnings growth and resilient demand could support DDOG, SNOW, RH, WSM, ETH, LZB, LESL and TPX over the next year. The evidence is mixed. DDOG has the clearest current operating support. WSM has demonstrated share and execution in a difficult category. RH’s premium positioning may help it take share, but the housing backdrop remains a constraint. The rest of the group needs a more visible volume recovery, better consumer confidence, or both.

Software: demand is visible, but the bar is high

DDOG’s latest reported quarter is the strongest evidence in the group. Datadog reported Q2 revenue of $1.12 billion, up 36% year over year, with roughly 4,720 customers above $100,000 in annual recurring revenue versus about 3,850 a year earlier. Management said growth among non-AI customers accelerated to the high 20s and described demand as healthy across startups and large enterprises.^[2]

That is more than an AI slogan: it links customer usage, cloud migration and observability demand. But it also raises the standard the market may apply to the stock. A company can deliver strong growth and still face valuation or expectation risk when the bar is elevated. Current news coverage highlighted DDOG’s five-day advance and the debate over the price investors are paying for that growth.^[3]

SNOW presents a different kind of test. News reports said the company proposed a $3.5 billion convertible-notes offering, with shares pressured by dilution concerns. That financing event is not evidence that product demand has weakened, but it can change the near-term way investors value growth and capital needs.^[3] The lesson for the thesis is straightforward: resilient usage helps, but shareholder returns also depend on financing structure, dilution and the price of growth.

Home furnishings: execution is real, recovery is not yet broad

The consumer side contains encouraging pockets, but the transcript record does not show a synchronized recovery. Williams-Sonoma management reported positive comps in both furniture and non-furniture categories in its latest available call, with e-commerce up 4.8% and retail up 4.7%. It also said the company gained share while maintaining full-price selling. At the same time, gross margin was 44%, down about 30 basis points, and merchandise margin fell 100 basis points as tariffs and fuel costs flowed through.^[4]

That combination matters. WSM is evidence that brand breadth, channel execution and full-price discipline can offset a weak category. It is not proof that industry-wide unit demand has recovered.

The broader housing-sensitive transcript evidence is more cautious. One furniture manufacturer described demand as below desired levels and tied a stronger recovery to sustained home sales. Another said the market remained difficult, with consumer confidence and housing affordability weighing on unit sales.^[4] RH’s own recent call was framed around demand outpacing parts of the home-furnishings industry, but with a meaningful cost to margin.^[4]

That leaves RH, WSM, ETH, LZB, LESL and TPX exposed to a common question: can company-specific execution keep earnings moving while the underlying market remains soft? The answer may differ by brand and balance sheet, but the macro input is shared.

Macro: not recessionary, but not easy

The latest macro snapshot available to this research pass shows real GDP growth of 2.1% year over year and unemployment at 4.1%, while the recession flag is false. Those are supportive conditions for a base case of continued economic activity.^[5]

The counterweights are equally important. CPI inflation was 3.35% year over year, the 10-year Treasury yield was 5.18%, and consumer sentiment was 51.7. The VIX was only 14.21, so the market was not pricing a generalized panic even as rate sensitivity and weak sentiment remained relevant.^[5]

For software, this backdrop can still work if customers see measurable returns from cloud, security and AI workloads. For furniture and bedding, the transmission mechanism is slower: housing affordability, mortgage rates, home transactions and confidence affect when consumers make large discretionary purchases. The thesis therefore has two different clocks. Cloud usage can compound inside existing customers; home-furnishings volume often needs a broader change in household behavior.

Evidence scorecard

This table is a research framework, not a ranking or a recommendation. The current evidence supports a selective version of the hypothesis, not the entire basket as one trade.

What to watch next

1. Usage quality in software. Watch whether DDOG and SNOW show broad-based customer expansion, not merely AI-related activity or a small number of large accounts.
2. Margins alongside demand. WSM’s latest result shows why revenue growth alone is insufficient: merchandise and gross-margin trends will determine how much demand converts into earnings.
3. Housing transmission. For RH, ETH, LZB, LESL and TPX, look for evidence in traffic, orders, units and conversion before treating a share-price move as a category recovery.
4. Rates and confidence. A 5.18% 10-year yield and 51.7 consumer sentiment reading leave the consumer thesis vulnerable even without a recession.^[5]
5. Financing and expectations. SNOW’s proposed convertible financing is a reminder that capital structure can dominate a good operating story over short intervals.^[3]

Bottom line

The cleanest conclusion is not that resilient demand has failed; it is that resilience is concentrated. DDOG offers the strongest direct evidence of enterprise usage growth in this group. WSM offers evidence of company-specific share and channel execution. The housing-sensitive names need a broader improvement in affordability, confidence and transaction activity that the current data does not yet establish.

For the next year, the bullish case requires two things to be true at once: software customers keep expanding workloads with measurable returns, and consumer-facing companies turn selective brand strength into volume without surrendering margins. The cautious case requires only that one of those links breaks. That asymmetry is why the tape is separating the names rather than confirming the hypothesis wholesale.