|
GTM stack intelligence, enriched.
|
|
Salesforce reported Q1 FY27 earnings on May 27th and their usage-based products showed true enterprise results. Agentforce generated $1.2 billion in ARR, up 205% year-over-year, with 3.8 billion “Agentic Work Units” delivered during the quarter. Salesforce has been a leader in the evolution of SaaS over the last two decades. That comes with scale, and with scale comes complexity and inertia. For anyone who was on the fence about enterprise level transitions to consumption-based pricing models, Salesforce just proved you can make the transformation at scale and see meaningful revenue generation.
In April, HubSpot provided the same confirmation from a different angle. Breeze’s outcome-based pricing, $0.50 per resolved conversation for the Customer Agent and $1 per qualified lead for the Prospecting Agent, has been running across 8,000 customers at a 65% resolution rate. Outcome-based pricing and usage-based pricing are two different structures, but they point towards the same direction.
The practical implication for your stack is becoming clearer; if you’re still budgeting for AI tools on per-seat assumptions, you’re modeling against a contract structure that is beginning to dissolve.
|
|
|
Salesforce — Agentforce Posts $1.2B ARR; Consumption Billing Is No Longer an Experiment
Salesforce’s Q1 FY27 earnings (May 27) delivered the most significant data point for GTM AI adoption this year. Agentforce hit $1.2 billion in ARR representing 205% year-over-year growth. Agentforce and Data360 processed 28.6 trillion tokens during the quarter, up 152% QoQ.
The consumption billing model Salesforce introduced with Agentforce could have been dismissed as a pilot structure when it was first announced. Now that its ARR has surpassed $1B, that position becomes less tenable. At 3.8 billion work units in a single quarter, the billing model is being stress-tested at production scale, and so far it’s passing. Salesforce is not pivoting back.
For RevOps teams currently renewing or evaluating Salesforce contracts: the SKU bookings number matters here. More than 50% of Agentforce and Data360 bookings came from existing customers, not just selling Agentforce to net-new logos. If you’re mid-renewal, expect Agentforce upsells as part of the conversation, with consumption billing as the pricing structure. If your team is new to that pricing structure, it is likely time to understand how that affects your business model.
Source: salesforce.com — Q1 FY27 earnings
|
|
|
Asana — $75M Acquisition of StackAI Signals Workflow Layer Play
Asana acquired StackAI on May 28th for $75 million. StackAI is a no-code agent builder with native connectors to Salesforce, HubSpot, and Slack. Asana’s framing in the announcement: “the operating system for human-agent teams.”
The “operating system” positioning is ambitious for a project management platform, but the logic is straightforward enough. Instead of re-architecting a new platform that can handle workflows across GTM and CS systems, Asana will sit above specific tools and manage how work moves between them. Layering in agent connectors provides a credible path towards evolving existing workflows into agentic systems, rather than building net new from the ground up. StackAI’s cofounder Tony Rosinol captured it perfectly: “AI creates ROI for enterprises when agents can specialize and reach into the systems where business actually runs. General-purpose agents talk; specialized agents act.”
The practical question for GTM teams will be where Asana fits in their stack relative to the growing capabilities of Salesforce Flows, HubSpot Workflows, and emerging MCP-connected agents. The key questions will be whether RevOps teams need to connect agentic workflows beyond GTM tools and into ERP and ITSM platforms, and whether Asana will continue to expand those connectors aggressively post-acquisition.
Source: asana.com/press · techcrunch.com
|
|
|
HubSpot — Smart CRM Index Beta and Breeze Workflow Code Generation
Two HubSpot product updates were released. On May 27th, the revamped Smart CRM Index entered public beta. On May 28th, the developer changelog added Breeze’s ability to write custom code workflow actions from plain language.
The Smart CRM Index is the more immediate operational item for RevOps teams. The indexing view for contacts, companies, deals, and custom objects is getting a simpler, cleaner UI. You can opt-in starting May 27, but every portal will be migrated to the new layout by July 31st. Begin preparing your teams for the upcoming change regardless of whether you plan to opt-in early or not.
The Breeze code generation change may be more consequential for teams that have been blocked by technical complexity in building custom workflow actions. For lean RevOps teams, writing automation logic in plain language closes a gap that previously required a HubSpot admin or professional services.
Source: community.hubspot.com · developers.hubspot.com
|
|
The tool-layer changes above are context for what follows.
|
|
|
Clay — Claygent Builder Ships a Native Agent Environment for GTM Teams
Clay shipped Claygent Builder on May 7th — a native environment for building, testing, and deploying GTM agents in natural language, available free on every plan including trials. Clay having agents is not the news; Claygents have been in the product for a while. It’s that the build-and-test loop previously required jumping between tabs, and every test run consumed credits. Claygent Builder consolidates that into a single interface where agents are built conversationally, tested free on real data, and deployed across tables without being rebuilt from scratch each time.
The reduced friction to utilize Claygents is also notable. My guess is that Clay realizes the barrier to building agentic workflows is shrinking with each new MCP release. Allowing users to build and test agents on real data for free on existing plans reduces that friction and keeps Clay as the tool of choice for GTM Engineers.
The practical readiness question is whether your team’s Clay usage is already at a depth where a native agent builder accelerates existing workflows, or whether you’d be learning the agent layer and the enrichment layer simultaneously.
Source: community.clay.com — Claygent Builder announcement
|
|
The rest of the week, in brief:
|
|
|
|
Microsoft Copilot Multi-Model: The May 2026 Copilot update ships GPT-5.5 as the default for most interactions and adds Claude Opus 4.8 as a selectable model inside Microsoft 365. A Finance Agent is in preview. The multi-model architecture confirms Microsoft’s approach: best-available model for the task type rather than single-vendor optimization.
|
|
Anthropic $65B Series H + Claude Opus 4.8: Anthropic closed a $65 billion Series H at a $965 billion post-money valuation (May 28). Claude Opus 4.8 shipped alongside the announcement. For GTM teams evaluating Anthropic-dependent vendors: API pricing decisions at this scale of capital deployment will eventually reflect the investment structure.
|
|
|
👁️ Noticed
Asana describes StackAI as giving them “an operating system for human-agent teams.” Salesforce has been calling Agentforce the “digital labor platform.” HubSpot’s framing is “the customer platform for the AI era.” Every vendor in the GTM stack is currently competing to be named the OS layer of AI-augmented GTM. Calling yourself the OS doesn’t make the others peripheral.
|
|
|
|
|
My read on the week:
3.8 billion is a big number. That’s the Agentic Work Units Salesforce delivered in a single quarter through Agentforce. Agentic Work Units are the building block for running and measuring agentic capabilities. The reason isn’t its size, but rather what it says about the billing model attached to it. Agentforce runs on consumption pricing. At 3.8 billion units, Salesforce has deployed that pricing structure at production scale, across thousands of enterprise customers. More importantly, it has generated material revenue announced on an earnings call.
HubSpot ran the same confirmation from a different direction. Their Breeze outcome-based pricing, half a dollar per resolved conversation and a dollar per qualified lead, has been running across 8,000 customers at a 65% resolution rate. While outcome-based and usage-based pricing are not the same, they both point towards the continued evolution away from pure seat-based pricing models at enterprise scale. A small collection of data companies began this transformation several years ago, but the adoption of a usage-based pricing model from the frontier AI labs has been the catalyst forcing every SaaS company to reconsider their pricing structure to stay compatible and competitive in this new paradigm.
This connects directly to Enriched’s framework: per-seat pricing structures are getting squeezed by in-house AI builds and vendors who can now credibly offer usage-based alternatives. The vendors repricing toward consumption aren’t doing it out of generosity, they’re doing it because the model provides the customer pricing flexibility while delivering longer-term revenue growth options for the vendor. The RevOps implication is that your budget modeling for CRMs, enrichment tools, access to frontier AI models, and any other vendor in your GTM stack needs to include consumption pricing structures. Just this past week Uber announced they consumed a full year of budgeted tokens in just 4 months. You need to estimate agent activity volume, not just licensed users.
Asana acquiring a no-code agent builder has a parallel logic; vendors are acknowledging that survival depends on being connectable to agentic workflows, not just interoperable with the CRM. The evaluation has shifted from “does your tool have AI features” to “can an AI agent take action inside your tool as part of a bigger workflow” and even more importantly “does the commercial structure scale with my usage, not just with our size.”
Ahead of your next renewal conversation, take the time to model the agent activity volume you expect to run over the next 12 months on Salesforce or HubSpot spend — would your budget go up or down relative to your current seats?
See you next week. — Andrew
|
|
|
|
|
|
