Title: PRO: This Week in Visuals (AVGO HPE VEEV IOT RBRK LULU GTLB) Show: App Economy Insights (How They Make Money) Author: Bertrand Hartman Date: 2026-06-06 URL: https://www.appeconomyinsights.com/p/pro-this-week-in-visuals-ede Note: PRO/paid Substack post, saved verbatim for personal study. Prose article β€” no (mm:ss) timestamps. Body reproduced as published. --- πŸ“Š PRO: This Week in Visuals (Jun 6, 2026). At a glance: Broadcom's AI bookings triple shipments; HPE's AI demand pulls forward; Veeva's Falcon targets agentic labor; Samsara scales operational AI; Rubrik becomes mission control for the AI era; Lululemon's turnaround stalls; GitLab pivots to AI with layoffs. 1. πŸ“‘ Broadcom: AI Bookings Triple Shipments β€” Q2 revenue +48% Y/Y to $22.2B ($70M beat), non-GAAP EPS $2.44 ($0.04 beat). Semiconductor solutions +79% to $15B, driven by AI semiconductor revenue +143% Y/Y to $10.8B. Infrastructure software +9% to $7.2B. Operating margin a record 49%; FCF $10.3B (46% of revenue). Shares fell ~15% post-earnings, giving back most of the $270B market cap added over the prior five sessions. Broadcom is seen as a beneficiary of Google's ~$85B equity raise to fund its AI buildout (Google is one of Broadcom's six core custom XPU customers). Headline metric was order velocity: AI semiconductor bookings exceeded $30B, ~3Γ— what it shipped, with CEO Hock Tan saying "our visibility runs all the way to 2028." AI networking ~40% of AI revenue (Tan expects normalization toward 30% as custom XPUs ramp). Expectations were sky-high: Q3 AI semi guide $16B (200% Y/Y) missed the $17.2B buy-side estimate; Tan implied FY26 AI revenue ~$56B vs $57.6B consensus; the $100B FY27 AI target was reiterated, not raised; multi-year commitments are back-end loaded into late FY27/FY28. Customer deployments: 1.3 GW for OpenAI in 2027 (part of a 10 GW deal through 2029); 3 GW for Meta through end-2028 (first 1 GW in H2 2027); a $35B debt financing with Apollo and Blackstone to fund Anthropic's purchase of Google TPU chips Broadcom helped develop, Broadcom backstopping the largest portions β€” part of a broader AI XPU Platform targeting 20+ GW of LLM compute through 2028. Guided Q3 revenue $29.4B (vs $28.47B consensus), semis $20.5B (+124% Y/Y). The sell-off was about failing to raise the $100B FY27 AI target + Tan acknowledging Google will diversify its TPU supply. 2. πŸ–₯️ HPE: AI Demand Pulls Forward β€” Q2 revenue +40% Y/Y to $10.7B ($0.9B beat), non-GAAP EPS more than doubled to $0.79 ($0.26 beat). Gross margin 37%, operating margin 7%. Shares jumped as much as 37% (biggest gain since the 2015 HP split); stock nearly doubled over the past month. Networking +148% Y/Y to $2.7B on the Juniper acquisition (data-center networking >tripled to $320M; routing $775M from $1M a year ago). Cloud & AI +23% to $7.7B; server +33% to $5.5B on triple-digit traditional server booking growth. AI systems orders +$1.8B in the quarter (cumulative AI bookings $16.4B); entered Q3 with a $5.9B backlog skewed to enterprise/sovereign. CEO Antonio Neri: "a volume story with very disciplined pricing execution," but supply is the challenge β€” DRAM and NAND pricing pressure could persist into 2027. Elliott's Chris Hsu joined the board. Raised FY26 revenue-growth guide to 29–33% (from 17–22%), EPS ~$3.40 (vs $2.43 consensus), FCF to β‰₯$3.5B (from $2B); introduced an FY27 framework of 8–12% growth and β‰₯$4.5B FCF β€” pulling long-term targets in by two years. 3. πŸ’Š Veeva: Falcon Targets Agentic Labor β€” Q1 revenue +16% Y/Y to $883M ($25M beat), non-GAAP EPS $2.24 ($0.10 beat). Subscription +15% to $730M; record services quarter. Launched Veeva Falcon β€” a platform built around agentic labor, AI agents designed to fully replace specific human job functions in life sciences (clinical-trial documentation, safety case triage), running "headless" against Veeva's systems of record. Rolling out under a "models, agents, and applications" (MAP) architecture for the last mile of regulated workflows. Vault CRM crossed 150 live customers (from 125) with an 80% win rate. The Ostro acquisition closed (~$10M revenue over three quarters). Uses Claude Code internally; will hire less as AI productivity scales. Raised FY27 revenue guide to $3.635–3.645B and EPS ~$9.05; reiterated the path to $6B revenue by 2030. 4. πŸš› Samsara: Operational AI Scales β€” Q1 revenue +31% Y/Y to $479M ($24M beat), non-GAAP EPS $0.17 ($0.04 beat). ARR crossed $2B (+30% Y/Y); third consecutive GAAP-profitable quarter (helped by a $30M arbitration award vs Motive). Shares βˆ’3% after hours as Q2 guide landed in-line. $100K+ ARR +37% Y/Y (96% on 2+ products); $1M+ ARR +62%. Emerging products (Connected Asset Maintenance, AI offerings) >20% of net new ACV. New: Waste Intelligence and Ground Intelligence (frontline labor automation). CEO Sanjit Biswas: "bits to atoms." Q2 guide $482–484M implies 23–24% growth (vs 31% in Q1). Raised FY27 revenue guide ~$39M to $2.01B (24% growth), EPS $0.71, 20% non-GAAP op margin target. 5. πŸ›‘οΈ Rubrik: Mission Control for the AI Era β€” Q1 revenue +39% Y/Y to $387M ($21M beat), non-GAAP EPS $0.16 ($0.19 beat). Subscription +41% to $374M; subscription ARR +32% to $1.57B; NRR ~120%; gross margin 81%; FCF margin 19%. Shares βˆ’2% after hours. Identity protection crossed $50M ARR (+38% sequential); $100K+ ARR customers +24% to 2,946. Rubrik Agent Cloud GA with the Semantic AI Governance Engine (SAGE) β€” real-time agent guardrails + "Agent Rewind." Project Glasswing: joined Anthropic's program for early access to Mythos research. The American Hospital Association selected Rubrik as a preferred cybersecurity provider (~5,000 member hospitals). CEO Bipul Sinha held 50% more CIO/CISO meetings sequentially (prevention-first β†’ recovery-first vs agentic AI threats). Subscription ARR growth decelerated 34%β†’32%. Raised FY27 revenue guide $39M to ~$1.643B; Q2 guide $396M (vs $383M consensus). 6. 🧘 Lululemon: Turnaround Stalls β€” Q1 revenue +4% Y/Y to $2.5B ($40M beat), GAAP EPS $1.69 ($0.02 beat), down from $2.60. Operating income βˆ’37%; operating margin 11% (from 18%). Shares βˆ’11% post-earnings; βˆ’40% YTD (one of the worst in the S&P 500). Americas revenue βˆ’3% (comps βˆ’6%); China Mainland +30% (comps +13%, CNY timing +8pts); Rest of World +13%. Gross margin βˆ’4pp to 54% (3pp tariffs, 1pp deleverage). CFO Meghan Frank: the brand "took a beating in the media and on social channels." Settled the proxy fight with founder Chip Wilson (two new directors + a third with apparel expertise, 18-month standstill; $11M battle cost behind it). Former Nike exec Heidi O'Neill becomes CEO in September (Frank + COO AndrΓ© Maestrini interim). Newness penetration ~30% (from 23%, target 35%). Cut FY26 revenue guide $35M to $11.0–11.15B (1% decline to flat), EPS to ~$11.05 (from ~$12.20); Q2 guide $2.45–2.475B (vs $2.59B consensus), op margin ~11.6% vs 20.7% a year ago. 7. 🦊 GitLab: AI Pivot With Layoffs β€” Q1 revenue +23% Y/Y to $264M ($10M beat), non-GAAP EPS $0.23 ($0.03 beat); operating loss narrowed to $16M (from $34M). Big news: a 14% workforce reduction (350 employees, exiting 22 countries, footprint βˆ’37%), $30–35M restructuring (mostly Q2); customers in exited markets served via partners. CEO Bill Staples: evolving "to be the trusted enterprise platform for software creation in the AI era." Duo Agent Platform (DAP) showing "promising traction" but CFO Jessica Ross flags it won't meaningfully contribute to FY27 revenue; expanded Google Cloud partnership lets Vertex AI users access DAP for agentic coding. Raised the low end of FY27 revenue guide to $1.112–1.118B and adjusted EPS to $0.79–0.82. Question: can the leaner cost structure fund enough DAP go-to-market to reaccelerate into FY28, and does AI-code-generation displacement risk keep the multiple compressed. Disclosure (footer only, not discussed in body): author owns CRM, SNOW, ZS, MDB, PATH, ESTC.