EWP · Pulse 2 of 6 · Agentic Demand Family · weekday composite
Enterprise Workflow Pulse
Where enterprise agents are actually being deployed: Agentforce AWUs in Salesforce, Now Assist customers at ServiceNow, Copilot seats at Microsoft, MCP server count across the developer ecosystem, Datadog's LLM observability span volume. The Pulse that decides whether agentic SaaS is real work or just demos. Six modules, weekday composite, equity-gap signal against CRM/NOW/MSFT/DDOG/SNOW/MDB/NET/TEAM.
59 points · since 2026-05-27
Module breakdown
| Module | Weight | Raw signal | Score | Status |
|---|---|---|---|---|
| M1 · Salesforce Agentforce | 25% | signal 95/100 | 95 | live · 2026Q3 |
Evidence — CRM 95/100 CRM 2026Q3 mass · accelerating confidence: high
“Agent force and data reached nearly $1.4 billion in ARR in the quarter, up a 114% year over year, including Agent Force ARR of about $540 million, 330% year over year...we've closed over 18,500 AgentForce deals. 9,500 of them are paid trans” Extracted from company earnings-call transcripts (AlphaVantage) via Claude Sonnet, temperature 0. Diary entry on our process, not investment advice. | ||||
| M2 · Copilot / Claude Code (dev adoption) | 20% | 15,100,156 npm/wk | 100 | partial — GitHub Copilot org-count next |
| M3 · ServiceNow Now Assist | 15% | — | 50 (neutral) | pending — awaiting transcript |
| M4 · Datadog LLM Obs + AI-native | 15% | — | 50 (neutral) | pending — awaiting transcript |
| M5 · MCP ecosystem (GitHub + npm) | 20% | 51,726,406 weekly downloads | 100 | live |
| M6 · Snowflake + MongoDB (RAG/data) | 5% | signal 67/100 | 65.2 | live · 2026Q2 |
Evidence — SNOW 72/100 · MDB 62/100 SNOW 2026Q2 scaling · accelerating confidence: high
“AI is a core reason why customers are choosing Snowflake, influencing nearly 50% of new logos won in Q2...powering 25% of all deployed use cases with over 6,100 accounts using Snowflake's AI every week.” MDB 2026Q2 early · accelerating confidence: medium
“a leading electric vehicle company chose Atlas and Vector Search to power its autonomous driving platform...they selected MongoDB for superior performance at scale...now rely on Atlas to handle over 1 billion vectors and expect 10 times gro” Extracted from company earnings-call transcripts (AlphaVantage) via Claude Sonnet, temperature 0. Diary entry on our process, not investment advice. | ||||
M5 (MCP ecosystem — GitHub stars + npm weekly downloads) and M2 (Copilot / Claude Code dev adoption) are live lead modules, re-fetched each run. Substance modules M1/M3/M4/M6 are scored from the latest company earnings-call transcripts via the Earnings-NLP pipeline (AlphaVantage transcript → Claude Sonnet extraction, temperature 0, confidence-damped), cached per fiscal quarter — expand “Evidence” on each to see the underlying metrics and the verbatim quote. Modules awaiting a fresh transcript hold a neutral 50 until the next quarter lands.
Multi-Basket Equity Gaps v3 · value-capture decomposition
Three sub-baskets organized by where in the enterprise-workflow stack value is being captured: Workflow Platforms (CRM/NOW/MSFT/TEAM) where agents are authored, Observability + Governance (DDOG/NET) where they're monitored, and the Data + RAG Layer (SNOW/MDB) they call against. The cross-read is what reveals rotation — when one layer accelerates while another stalls, the market is pricing-in a stack-shift.
| Sub-Basket | Vertical | Weight | 21d Basket Return | Module Score | Tickers |
|---|---|---|---|---|---|
| Workflow Platforms | Where agentic workflows actually get authored + run | 45% | +33.00% | 100.0 | CRM · NOW · MSFT · TEAM |
| Observability + Governance | How agentic workloads are monitored + routed | 25% | +6.64% | 83.2 | DDOG · NET |
| Data + RAG Layer | Vector + warehouse layer agents call against | 30% | +30.95% | 100.0 | SNOW · MDB |
Methodology
EWP is Pulse 2 of 6 in the Closelook Agentic Demand Family. Pulse weight philosophy per Phase 2 spec: M5 MCP ecosystem (20%) and M1 Agentforce (25%) are the highest-conviction anchors — MCP ecosystem because adoption-numbers (5-6M weekly npm downloads, 86k+ GitHub stars) signal mainstream production, not developer experiment; Agentforce because CRM disclosed $800M ARR + 2.4B AWUs + 169% YoY at Q4 FY26 — T3 mass-production maturity per the family-wide Maturity-Score sub-module.
Weekday-only composite (06:00 UTC). Lead modules (M2, M5) re-fetch each run; substance modules (M1, M3, M4, M6) re-evaluate quarterly with each earnings cycle. Equity Gap activates once ~3 months of EWP composite history accumulates.
Pulse reading is a diary entry on our process, not investment advice.
FAQ · from the current data · as of 2026-08-16
Quick answers
What does the Enterprise Workflow Pulse measure?
EWP reads where enterprise agents are actually being deployed — Agentforce AWUs at Salesforce, Now Assist customers at ServiceNow, Copilot seats at Microsoft, MCP server count across the developer ecosystem, and Datadog LLM observability span volume — into a weekday composite.
What is the current EWP composite reading?
As of Aug 16, 2026, 08:00 AM GMT+2, the EWP composite reads 82 out of 100, a Surging momentum regime, with 4 of 6 modules currently live.
Which stocks make up the EWP equity basket?
CRM, NOW, MSFT, DDOG, SNOW, MDB, NET and TEAM — the workflow-platform, observability and data/RAG-layer names most exposed to enterprise SaaS-agent adoption.
How are the six EWP modules weighted?
Salesforce Agentforce carries 25%, the MCP ecosystem (GitHub + npm) 20%, Copilot/Claude Code dev adoption 20%, ServiceNow Now Assist 15%, Datadog LLM observability 15%, and Snowflake/MongoDB RAG 5%.