Forge is chasing real, urgent demand — but the sale will be hard-won, not easy
Asked (summary):
Given Forge's positioning as a control plane for enterprise AI agents — discovery, identity, MCP/tool governance, behavioral baselining, runtime enforcement — and its claimed early traction, will it be hard for Forge to do well and sell the product easily?
This assessment draws on 15 one-per-vendor competitive profiles (from 13 independent hosts) and 10 sourced market-demand findings (from 10 hosts), gathered as of 3 November 2026. The competitor list is illustrative, not exhaustive; demand metrics come from different surveys with different populations and definitions, several via secondary write-ups of primary research, and should not be combined as one survey. Net read: medium-high go-to-market difficulty (~7/10) with a credible, execution-sensitive opportunity.
Two pressures at once: strong market pull, crowded field pushing back
Left: how urgent the problem is — reported % per finding (each from a distinct survey; do not sum). Right: 15 overlapping vendors in four camps; dark marks already own an enterprise control point or installed base. Hover any mark for detail.
Market pull — reported % from one survey/reportCompetitor that owns an enterprise control point / installed baseCompetitor without that installed-base advantage
Pilot-to-production gaps recur across independent surveys — 85% vs 5% (Cisco), 62% vs 23% (McKinsey), 88% of POCs never shipping (IDC) — and each names governance, not model capability, as the blocker. siroccogroup.com
Only 14.4% of organizations report their agent fleet went live with full security and IT approval, and 57.4% of builders cite missing logging and audit trails — the visibility gap Forge’s discovery and audit story targets. gravitee.io
Seven of the 15 profiled competitors already own an enterprise control point or installed base — Microsoft, Google, Okta, CrowdStrike, Snowflake, Boomi, and Palo Alto via its Portkey acquisition — making “wait for the incumbent” a live objection. arcjet.com
But incumbent coverage is leaky today: Microsoft Agent 365 is in public preview with governance split across three admin surfaces, Okta’s Agent Gateway is a research release, and CrowdStrike’s MCP proxy excludes non-MCP traffic — the window Forge must convert before it closes. paulbien.com
The commercial answer
Will Forge find customers? Likely yes — among enterprises already pushing agents into real workflows, especially regulated, multi-platform environments. Will it sell easily? No. Expect technical, consultative, founder-assisted enterprise selling, not a transactional motion. Treat the company-supplied figures — 8 active deployments, ~20 POVs expected in two months, a 100% head-to-head win rate, a $10M seed — as unverified claims; a 100% win rate is not decision-useful without the denominator, the named competitors, the definition of a win, and whether signed production contracts resulted.
Why the sale is hard
Category and ownership ambiguity: CISO, IAM, AppSec, AI platform and infrastructure all plausibly own the budget
Buyers can wait for Okta, Microsoft, Palo Alto (Portkey/Prisma AIRS), or CrowdStrike, or take bundled capabilities
A new inline control plane draws latency, reliability, false-positive and trust objections; enforcement must be proven in production
Agent deployments remain early, and security ROI must be shown before a visible incident
Why Forge can win
Vendor-neutral coverage across identity + MCP/tools + runtime, where incumbents are siloed, ecosystem-bound, preview-stage, or proxy/MCP-only
Deployment through existing EDR, network, IdP and gateway control points
Per-agent × human behavioral baselines and a sub-50ms inline claim
Best wedge: regulated enterprises with production-bound, multi-ecosystem agents, a named executive security owner, and one urgent control problem — least privilege, approval-gating money-moving actions, MCP/tool risk, shadow-agent discovery, or audit evidence
GTM recommendation
Sell one painful production outcome, not "the control plane for everything"
Land with a 30–60 day production POV with pre-agreed success metrics
Quantify: time to inventory, unsafe actions blocked, excess permissions removed, approval latency, false positives, deployment effort
Expand from one agent workflow or business unit
Diligence questions for the Founding AE
Exact denominator behind the 100% win rate; which competitors appeared in live bake-offs
How many of the 8 deployments are paid production customers; ACV and contract length; POV-to-paid conversion
Security buyer and budget source; sales-cycle length; referenceability; deployment dependencies
Gross retention/expansion; how much selling still depends on the CEO
Verdict: Forge has a plausible path to becoming important, but the Founding AE is being hired to create a repeatable category sale — not inherit an easy one.
The evidence, row by row
10 demand findings, then 15 competitive profiles. Each row links its source.
Type
Who
What the row says
Implication / opening for Forge
Source
Method: 25 rows total — 15 one-row-per-vendor competitive profiles (13 independent hosts) grouped into direct agent-security specialists, identity/IAM, major security & cloud platforms, and MCP/gateway infrastructure; and 10 market-demand findings (10 hosts), each a percentage from a distinct survey or report with its own population and definitions — not combinable. Several demand rows are secondary write-ups citing primary research. No single URL backs more than 10% of either dataset. Research date 2026-11-03. Company traction figures (8 deployments, ~20 POVs, 100% win rate, $10M seed) are unverified claims. Long capability lists trimmed in tooltips for space; full text in the table.