Software Is Passing the Demand Test; Home Names Still Need Proof

The supplied hypothesis is that earnings growth and resilient demand can support DDOG, SNOW, RH, WSM, ETH, LZB, LESL and TPX over the next year. The opening evidence is more selective: software demand has a credible operating signal, while home-furnishings demand is improving at some brands but remains exposed to housing, rates, tariffs and consumer confidence.

The tape is asking for proof, not a slogan

At 12:08 p.m. ET on September 23, the broad tape was softer: SPY was down 0.62%, QQQ 0.91%, XLY 1.41% and XLK 0.70%. Yet IGV, the software ETF, was up 1.38%. That relative performance is not proof of durable earnings growth, but it is a useful market test: investors were distinguishing software exposure from the broader growth and discretionary complex rather than selling every risk asset indiscriminately. The quote feed is 15 minutes delayed and marked regular-session data.^[1]

Within the specified names, DDOG was up 2.63% and SNOW 0.66%, while RH fell 3.80%, WSM 1.48%, LZB 2.01% and LESL 0.50%. TPX’s quote in the same feed is stale—dated February 26, 2025—so it is excluded from today’s price comparison rather than treated as current. ETH was down 3.14%; the symbol is reported exactly as supplied and should not be conflated with the cryptocurrency ether.^[2]

Working thesis: the market is currently rewarding evidence of recurring, mission-critical software usage more readily than a generalized consumer-recovery narrative. That can change, but the burden of proof is different across the two groups.

Software: the strongest evidence is usage, not the AI label

Datadog’s recent call commentary provides the clearest operating evidence in this basket. In Q2 FY2026, management said more than 750 AI customers used Datadog and that all 10 of the leading AI companies were customers. It also reported that MCP tool calls had quadrupled quarter over quarter and were more than 22 times the Q4 FY2025 level. Those metrics connect AI activity to cloud consumption, observability and security workloads—the mechanisms through which demand can become revenue—rather than relying only on a broad AI narrative.^[3]

There is still a qualification. Datadog also disclosed a large customer renewal with a user reduction beginning in Q3, which management said was incorporated into guidance. That is a reminder that strong platform adoption can coexist with account-level optimization. The key question is whether expansion across products and customers outweighs such usage pressure.^[3]

SNOW belongs in the same demand conversation, but this pass did not retrieve a comparable transcript block with enough specificity to make a new operating claim about its latest quarter. Its stock was modestly positive at the time of the snapshot, but price action alone cannot establish earnings durability.^[2]

Home furnishings: resilience exists, but it is not uniform

Williams-Sonoma offers the most constructive evidence in the home group. In its Q2 FY2026 commentary, management said both furniture and non-furniture posted positive comparable sales, with e-commerce up 6.5% and retail up 5.5%. It also said the home-furnishings industry was essentially flat, implying that the company’s reported growth was primarily share gain. At the same time, gross margin was pressured by tariffs, and management described the housing market as stagnant.^[4]

That combination matters. WSM’s result supports a narrower claim—brand, assortment, channel execution and market-share gains can offset a weak category for a time—not the stronger claim that housing-sensitive demand has broadly recovered. The same call said tariff pressure had peaked in Q2 and should moderate, but that remains a company outlook rather than a macro guarantee.^[4]

The weaker tape in RH, WSM and LZB on September 23 suggests the market is not giving all home names the same credit. RH’s decline was the sharpest among the quoted home-furnishing names, while WSM’s operating evidence was better than its intraday price. That mismatch is worth monitoring: it may reflect valuation, expectations, rates or company-specific news, but the available data do not identify one cause with confidence.^[2]

Macro backdrop: resilient growth, uncomfortable rates

The latest macro snapshot available in this pass is through August 2026. Unemployment was 4.1%, real GDP growth was 2.1% year over year and industrial production was up 1.42% year over year. Those readings are consistent with an economy still expanding rather than an outright recession. But CPI inflation was 3.35%, the 10-year Treasury yield was 4.96%, consumer sentiment was 55.2, and the high-yield spread was 2.66%. The yield curve remained positive at 0.25 percentage point, while the VIX was 14.81.^[5]

For software, firm long-term demand can help absorb a higher discount rate when usage is measurable and products are embedded in critical workflows. For home furnishings, nearly 5% long-term Treasury yields and low sentiment create a more demanding environment for big-ticket purchases, even if aggregate retail demand has held up. Reuters reported that August retail sales were robust but also noted building inflation pressure, reinforcing the idea that “resilient consumer” and “easy consumer conditions” are not synonyms.^[6]

Evidence scoreboard

What to watch next

1. DDOG’s November 5 report: the date is scheduled but marked estimated, before the open. The next update should test AI-customer growth, usage intensity, expansion and the disclosed renewal headwind.^[7]
2. SNOW’s December 2 report: scheduled but estimated, after the close. Watch consumption growth, remaining performance obligations and whether AI workloads are translating into durable platform usage.^[7]
3. WSM’s November 18 report: scheduled but estimated, before the open. The useful checks are category comps, market-share claims, full-price selling and tariff-related margin pressure.^[7]
4. The home group versus rates: RH, LZB and LESL need to show that demand can improve without relying on a broad housing rebound that has not yet appeared in the macro data.^[5]
5. Data quality before conclusions: TPX has no confirmed earnings date in the current calendar and its quote snapshot was stale. It should remain an unresolved item, not a positive or negative datapoint.^[7][2]

The balanced conclusion is that the hypothesis has a credible first leg and a conditional second leg. DDOG provides the clearest evidence that AI-related activity is feeding mission-critical software usage. WSM provides evidence that a strong operator can gain share in a flat home category. But the basket as a whole still faces high rates, inflation, tariff exposure, uneven consumer confidence and incomplete company-level evidence. The next earnings cycle—not today’s opening snapshot—will determine whether those isolated signals broaden.