ProgrammableMarketing/keycard.ai (opens in a new tab)
PDF

Views

  • Plan
  • Roadmap
  • Research

On this page

  • The brief
  • Start here
  • Idea portfolio
  • Effort vs impact
  • This week · 7
  • This quarter · 9
  • Big bets · 3
  • Wildcards · 3
  • How we can help
Generated Oct 26, 2026
  1. Programmable
  2. Playbooks
  3. Keycard
Keycard logo

growth playbook · Oct 26, 2026

keycard.ai (opens in a new tab)

Keycard growth playbook

Diagnosis

Keycard has raised $38 million, has a founding team from Auth0, Okta and Snyk, ships fast, and has a strong voice in its category through the Insecure Agents podcast, Agent Baseline and developer events.

Its hiring shows the next step: a developer self-serve motion alongside top-down enterprise sales, with a Growth & DX PM to productize evaluations, a RevOps lead to bring GTM systems in-house, a revenue marketer for pipeline and a staff PMM for enterprise positioning.

The biggest gap is the front door: the free plan and the 20-minute quickstart both start with an access request. This playbook opens instant Starter access and instruments the evaluation first, adds proof (a Chime case study, comparison and use-case pages, a template gallery), fast event follow-up and PQLs, then bets on a secrets scanner, an Agent Baseline self-assessment and marketplace buying.

Growth Score

7/10

Biggest Opportunity

Letting developers start for free in minutes and turning their usage into enterprise pipeline.

Biggest Risk

A crowded, fast-moving category (Aembit, Descope, Arcade, Astrix and the large identity vendors) where rivals with instant sign-up can win developers before Keycard's gate opens.

01

Start here

The three moves to make first. Each one is written up in full below.

  1. 1

    Make the free Starter plan start instantly, so the 20-minute quickstart is really 20 minutes

    Measure the median time from access request to first login today.

    High impact First result in 1 month
  2. 2

    Instrument every quickstart step, so the team sees where evaluators stall

    List the five milestones and where each event fires.

    High impact First result in 2 weeks
  3. 3

    Turn the Chime quote into the first full case study

    Ask Chime for a 30-minute interview.

    High impact First result in 1 month

02

Your idea portfolio

22 ideas, grouped by when to run them. Open any idea for how to run it: the steps, who to approach, copy you can use, and how to measure it.

  • 7 ideas are marked Unexpected: plays you are unlikely to have tried.

Every idea placed by the effort it takes against the impact it could have. Quick wins sit top left; select a number to open that idea. The filters below narrow the list by status, timing, category and what Programmable can build.

Effort vs impactEach number is an idea. Select one to open it.
Impact →
Quick wins · 1Big bets · 11Fill-ins · 7Reconsider · 3
Effort →

Showing 22 of 22 ideas

This week 7

Quick wins with a clear first action.

  • 01

    Keycard's pricing page offers a free Starter plan, but its "Get started" button opens a form to "request a Keycard account" that ends with "We'll be in touch", and the docs quickstart (about 20 minutes, 45 lines of Python) lists that request as its first prerequisite. A developer who arrives ready to try it waits instead, while Aembit, Descope and Arcade all offer an instant tenant or sign-up. The new Growth & DX PM is hired to cut time-to-first-credential-exchange, and the RevOps lead to wire a self-serve motion: the gate is the first thing both will measure.

    High impactMedium effort First result in 1 month Product Manager

    First action

    Measure the median time from access request to first login today.

    How to measure it

    Median sign-up to first credential exchange

    Now
    –
    Target
    Under 30 minutes
    By
    1 quarter

    Steps

    1. 1

      Baseline

      Requests, approval time and first-login rate for the last 90 days.

      Product Manager, Growth & Developer Experience 2 days

    2. 2

      Set the rules

      Which sign-ups get instant access and which are reviewed.

      CEO 2 days

    3. 3

      Build instant org creation

      Work-email sign-up creates a Starter org with Admin Role.

      Software Engineer, Systems 2 weeks

    4. 4

      Rewrite the button

      "Start free" on pricing and docs; keep "Book a demo" for Enterprise.

      Product Manager, Growth & Developer Experience 1 day

    5. 5

      Prefill the quickstart

      Org callback URL and next step shown on first login.

      Product Manager, Growth & Developer Experience 1 week

    6. 6

      Route to HubSpot

      Sign-up, source and company to the right rep.

      Founding Revenue Operations Lead 1 week

    7. 7

      Watch abuse

      Weekly review of flagged sign-ups.

      Field engineer weekly

    8. 8

      Report

      Activation and sign-up-to-pipeline weekly.

      Product Manager, Growth & Developer Experience weekly

    First result
    1 month
    Cost
    Engineering time
    Owner
    Product Manager, Growth & Developer Experience

    Smallest test

    Give instant access to one week of work-email requests and compare their activation.

    Why it works

    Let work-email sign-ups on Starter create an organization in the console immediately (the 5,000-transaction hard cap already limits risk), land them in the quickstart with their callback URL prefilled, and keep a review queue only for signals that need it (free-mail domains, unusual volume). Send each new sign-up to HubSpot with its source so the field team can still reach out to large companies within a day.

    Every hour between interest and first credential exchange loses evaluators, and a self-serve start lets field engineers spend time on enterprise evaluations instead of approvals.

    Ready to use

    ChecklistInstant Starter launch list

    [ ] Work-email sign-up creates a Starter org
    [ ] Admin Role granted to the first user
    [ ] Quickstart opens with the org callback URL
    [ ] "Request access" copy removed from pricing and quickstart
    [ ] Sign-up and "How did you hear" source sent to HubSpot
    [ ] Target-account sign-ups alert a rep
    [ ] Flagged sign-ups reviewed weekly
    [ ] Activation dashboard: sign-up, first resource, first credential exchange

    Risks and how to handle them

    • Low-quality sign-ups

      The Starter hard cap and work-email rule limit exposure; review flagged ones.

    • Enterprise prospects skip the sales team

      Alert reps on target-account sign-ups within an hour.

    Evidence

    • Your websitePricing page: Starter "Free", "5,000 transactions/mo (hard cap)", "Unlimited users, agents & apps", "Community support"; Team "$500 /mo" with 100,000 transactions and "$1 per 1,000 additional"; Enterprise on annual commitment. Every plan's "Get started" button opens the same early-access form: "Sign up with your work email to request a Keycard account", ending "We'll be in touch". The form asks "How did you hear about us?" with options including "Podcast", "Event" and "AI agent". The site links no community forum or chat for Starter's community support (footer: X, GitHub, LinkedIn, Insecure Agents).Source
    • Your docsDocs quickstart: "Build an MCP tool that calls GitHub as the signed-in User ... About 20 minutes", "the server itself is about 45 lines of Python". Its first prerequisite: "A Keycard account with the Admin Role ... Keycard is in Early Access; you can request access here."Source
    • HiringProduct Manager, Growth & Developer Experience (remote US, posted August 31, 2026): "Every customer starts with an evaluation, and today a field engineer runs each one by hand: setup, first credential exchange, integration questions. Your job is to turn that walkthrough into product, so an evaluation can run well without us in the room." The PM will "Drive time-to-first-credential-exchange down", "instrument the journey so we can see where evaluators stall", and ship fixes that are often "a better quickstart, a template repo, an error message that says what to do next".Source
    • HiringFounding Revenue Operations Lead (posted October 8, 2026): "We sell top-down to large enterprises and we're launching a developer self-serve motion. Today, an agency manages our GTM systems." Takes over HubSpot and Clay, defines "lead and account scoring, routing, and attribution across sales-led and self-serve paths", automates "product-qualified lead alerts" and "event lead processing from list to rep assignment ahead of our next major events", and makes sure "trial signups and usage signals reach the right rep with context".Source
    • Competitor researchRivals' homepages: Aembit, "Identity Control Plane for AI Agents", with a "Get an Aembit Tenant" button; Descope, "Customer and Agentic Identity Platform" with "Agentic Identity Hub 3.0" and a "Sign up" button; Arcade, "The Actions Runtime for Enterprise AI Agents", with "Start Building"; Astrix, discovery and security for "AI agents & NHIs"; Oasis, an NHI platform now "acquired by Cyera".Source
  • 02

    The PM's brief says "If the data you need doesn't exist, you get it added before forming theories". Today evaluations are run by a field engineer, so drop-off lives in people's heads, not in analytics.

    High impactMedium effort First result in 2 weeks Product Manager

    First action

    List the five milestones and where each event fires.

    How to measure it

    Sign-ups reaching first exchange

    Now
    –
    Target
    50%
    By
    1 quarter

    Steps

    1. 1

      Define milestones

      Five steps everyone reports on.

      Product Manager, Growth & Developer Experience 2 days

    2. 2

      Add events

      Console and SDK events.

      Software Engineer, Data 1 week

    3. 3

      Build funnel

      By source and company.

      Software Engineer, Data 1 week

    4. 4

      Sync to CRM

      Milestones on the HubSpot record.

      Founding Revenue Operations Lead 1 week

    5. 5

      Stall nudges

      One message per stuck step.

      Founding Revenue Marketing Lead 1 week

    6. 6

      Review

      Weekly funnel review.

      Product Manager, Growth & Developer Experience weekly

    First result
    2 weeks
    Cost
    Engineering time
    Owner
    Product Manager, Growth & Developer Experience

    Smallest test

    Track the quickstart by hand for the next ten evaluations.

    Why it works

    Define five milestones (sign-up, first resource, first application, first credential exchange, first policy change), fire an event for each from the console and SDKs, and show a funnel by source and company. Send stalled evaluators a short, specific nudge for the step they stopped on.

    A measured funnel tells product, engineering and go-to-market where to spend the next sprint.

    Risks and how to handle them

    • Privacy concerns from security buyers

      Document exactly what is collected.

    Evidence

    • HiringProduct Manager, Growth & Developer Experience (remote US, posted August 31, 2026): "Every customer starts with an evaluation, and today a field engineer runs each one by hand: setup, first credential exchange, integration questions. Your job is to turn that walkthrough into product, so an evaluation can run well without us in the room." The PM will "Drive time-to-first-credential-exchange down", "instrument the journey so we can see where evaluators stall", and ship fixes that are often "a better quickstart, a template repo, an error message that says what to do next".Source
    • HiringFounding Revenue Operations Lead (posted October 8, 2026): "We sell top-down to large enterprises and we're launching a developer self-serve motion. Today, an agency manages our GTM systems." Takes over HubSpot and Clay, defines "lead and account scoring, routing, and attribution across sales-led and self-serve paths", automates "product-qualified lead alerts" and "event lead processing from list to rep assignment ahead of our next major events", and makes sure "trial signups and usage signals reach the right rep with context".Source
    • Your docsDocs quickstart: "Build an MCP tool that calls GitHub as the signed-in User ... About 20 minutes", "the server itself is about 45 lines of Python". Its first prerequisite: "A Keycard account with the Admin Role ... Keycard is in Early Access; you can request access here."Source
  • 03

    The only customer proof on keycard.ai is one quote from Chime about running agents against production "in days", and the sitemap has no customer pages. The Staff PMM is asked to publish a first case study in 90 days, and enterprise security buyers want to see a peer.

    High impactLow effort First result in 1 month Staff Product Marketing Manag…

    First action

    Ask Chime for a 30-minute interview.

    How to measure it

    Opportunities using the story

    Now
    –
    Target
    50%
    By
    2 quarters

    Steps

    1. 1

      Ask

      Customer approval.

      Staff Product Marketing Manager 1 week

    2. 2

      Interview

      Security and engineering leads.

      Staff Product Marketing Manager 1 week

    3. 3

      Write

      Story with numbers.

      Staff Product Marketing Manager 1 week

    4. 4

      Review

      Customer legal.

      Staff Product Marketing Manager 2 weeks

    5. 5

      Publish

      Page and one-pager.

      Head of Marketing 1 day

    6. 6

      Enable

      Use in sales decks.

      Staff Product Marketing Manager ongoing

    First result
    1 month
    Cost
    Staff time only
    Owner
    Staff Product Marketing Manager

    Smallest test

    Draft the story from the existing quote and ask for one number.

    Why it works

    Interview the Chime team on the before (how agents got credentials), the rollout (days to production, resources connected) and the result, publish it with numbers, and make a one-page version for security reviews.

    One detailed story from a regulated fintech does more for enterprise deals than logos.

    Risks and how to handle them

    • Customer review delays

      Start with an anonymised version.

    Evidence

    • Your websiteHomepage: "Runtime auth for agents"; header buttons "Request access" and "Book a demo"; one customer quote, from Chime ("We had agents running against production systems in days"); SOC 2 Type II and a trust center; the Agent Baseline white paper "developed in partnership with Docker and Snyk"; quotes from Insecure Agents podcast guests at Cursor, Vercel, Mastercard and J.P. Morgan; SDKs for Python, TypeScript, Ruby and Go; federates Okta, Entra, Google and Auth0; a "Talk to our agent" link in the footer.Source
    • Your websiteSitemap (32 URLs, checked October 23, 2026): home, about, blog, careers, contact, events, Keycard for Code, brand, pricing and legal pages, plus 21 blog posts (launches for coding agents and multi-agent apps, Okta cross-app access support, the Smallstep device identity partnership, the Anchor and Runebook acquisitions, two "I built a ... agent" tutorials). No customer or case-study pages, no comparison pages and no use-case or industry pages.Source
    • HiringStaff Product Marketing Manager (posted October 8, 2026): "help Keycard win enterprise security buyers" (security architects, CISOs, platform leaders); run "win-loss analysis, track competitors across the agent security landscape"; build competitor comparisons and objection handling, "a launch-tiering framework and playbook", analyst briefings, and in the first 90 days "Publish your first customer case study". Bonus experience: "turning developer usage into enterprise deals".Source
  • 04

    The PMM is hired to track "competitors across the agent security landscape" and build competitor comparisons. Buyers comparing agent identity tools search by name, and keycard.ai has no comparison pages.

    High impactMedium effort First result in 1 month Staff Product Marketing Manag…

    First action

    Pick the two rivals that appear most in deals.

    How to measure it

    Comparison-page demos

    Now
    –
    Target
    20 a month
    By
    2 quarters

    Steps

    1. 1

      Pick rivals

      From win-loss notes.

      Staff Product Marketing Manager 2 days

    2. 2

      Check facts

      Current rival pages.

      Staff Product Marketing Manager 3 days

    3. 3

      Write pages

      Fair tables.

      Staff Product Marketing Manager 1 week

    4. 4

      Legal review

      Claims and names.

      Legal 1 week

    5. 5

      Publish

      With a start-free path.

      Head of Marketing 1 day

    6. 6

      Refresh

      Quarterly.

      Staff Product Marketing Manager quarterly

    First result
    1 month
    Cost
    Staff time only
    Owner
    Staff Product Marketing Manager

    Smallest test

    Publish one page and track visits.

    Why it works

    Publish one honest page per rival: what each is built for (workload IAM, customer identity, tool runtime, NHI discovery), where Keycard fits (runtime authorization on behalf of users), and how they work together. Keep the competitor comparison internal and the page factual.

    Comparison pages catch high-intent searches and teach AI assistants how Keycard differs.

    Risks and how to handle them

    • Category moves fast

      Date every page and review each quarter.

    Evidence

    • Competitor researchRivals' homepages: Aembit, "Identity Control Plane for AI Agents", with a "Get an Aembit Tenant" button; Descope, "Customer and Agentic Identity Platform" with "Agentic Identity Hub 3.0" and a "Sign up" button; Arcade, "The Actions Runtime for Enterprise AI Agents", with "Start Building"; Astrix, discovery and security for "AI agents & NHIs"; Oasis, an NHI platform now "acquired by Cyera".Source
    • HiringStaff Product Marketing Manager (posted October 8, 2026): "help Keycard win enterprise security buyers" (security architects, CISOs, platform leaders); run "win-loss analysis, track competitors across the agent security landscape"; build competitor comparisons and objection handling, "a launch-tiering framework and playbook", analyst briefings, and in the first 90 days "Publish your first customer case study". Bonus experience: "turning developer usage into enterprise deals".Source
    • Your websiteSitemap (32 URLs, checked October 23, 2026): home, about, blog, careers, contact, events, Keycard for Code, brand, pricing and legal pages, plus 21 blog posts (launches for coding agents and multi-agent apps, Okta cross-app access support, the Smallstep device identity partnership, the Anchor and Runebook acquisitions, two "I built a ... agent" tutorials). No customer or case-study pages, no comparison pages and no use-case or industry pages.Source
  • 05

    Keycard already maintains templates, a getting-started MCP server, a Slack agent tutorial, a Terraform provider and a GitHub Actions action on GitHub, but most repos have 0 or 1 stars and the site has no page that shows them. The PM role names "a template repo" as a common fix.

    Medium impactLow effort First result in 2 weeks DevRel

    First action

    List the five templates that run end to end today.

    How to measure it

    Template-started evaluations

    Now
    –
    Target
    30%
    By
    1 quarter

    Steps

    1. 1

      Audit repos

      Which run cleanly.

      DevRel 2 days

    2. 2

      Write cards

      Five templates.

      DevRel 2 days

    3. 3

      Build page

      Gallery on the site.

      Head of Marketing 1 week

    4. 4

      Add links

      Docs, pricing, blog.

      DevRel 1 day

    5. 5

      Track

      Clones and console sign-ups.

      Product Manager, Growth & Developer Experience weekly

    6. 6

      Add more

      One template a month.

      DevRel monthly

    First result
    2 weeks
    Cost
    Staff time only
    Owner
    DevRel

    Smallest test

    Publish a gallery with three templates.

    Why it works

    Publish a gallery page: one card per template with what it builds, time to run, language and a "run it" button that opens the console. Link it from docs, pricing and every blog tutorial, and ask workshop attendees to star the repos.

    Developers start from working code; a gallery turns repos nobody finds into an evaluation path.

    Risks and how to handle them

    • Templates break

      Run each in CI weekly.

    Evidence

    • GitHubgithub.com/keycardai: about 20 public repositories pushed in October 2026, including the TypeScript, Python, Go and Ruby SDKs, "templates" ("Keycard templates for building secure agents"), getting-started-mcp, a Terraform provider, a GitHub Actions auth action, a Slack agent tutorial and an egress proxy. The most-starred repository has 4 stars; most have 0 or 1.Source
    • HiringProduct Manager, Growth & Developer Experience (remote US, posted August 31, 2026): "Every customer starts with an evaluation, and today a field engineer runs each one by hand: setup, first credential exchange, integration questions. Your job is to turn that walkthrough into product, so an evaluation can run well without us in the room." The PM will "Drive time-to-first-credential-exchange down", "instrument the journey so we can see where evaluators stall", and ship fixes that are often "a better quickstart, a template repo, an error message that says what to do next".Source
    • Your docsDocs quickstart: "Build an MCP tool that calls GitHub as the signed-in User ... About 20 minutes", "the server itself is about 45 lines of Python". Its first prerequisite: "A Keycard account with the Admin Role ... Keycard is in Early Access; you can request access here."Source
    • Your websiteSitemap (32 URLs, checked October 23, 2026): home, about, blog, careers, contact, events, Keycard for Code, brand, pricing and legal pages, plus 21 blog posts (launches for coding agents and multi-agent apps, Okta cross-app access support, the Smallstep device identity partnership, the Anchor and Runebook acquisitions, two "I built a ... agent" tutorials). No customer or case-study pages, no comparison pages and no use-case or industry pages.Source
  • 06

    Keycard runs workshops, demo nights and panels across AI engineering and security events, and both the RevOps and Revenue Marketing posts ask for event follow-up "within days", automated "ahead of our next major events". Workshop attendees who built an MCP server with Keycard are the warmest leads it has.

    High impactMedium effort First result in 2 weeks Founding Revenue Operations L…

    First action

    Measure follow-up time for the last three events.

    How to measure it

    Event follow-up within 48h

    Now
    –
    Target
    95%
    By
    1 quarter

    Steps

    1. 1

      Measure

      Last three events.

      Founding Revenue Operations Lead 1 day

    2. 2

      Automate import

      List to HubSpot.

      Founding Revenue Operations Lead 1 week

    3. 3

      Enrich and score

      Clay and scoring.

      Founding Revenue Operations Lead 1 week

    4. 4

      Assign reps

      Routing rules.

      Founding Revenue Operations Lead 2 days

    5. 5

      Write follow-ups

      By event type.

      Founding Revenue Marketing Lead 2 days

    6. 6

      Report

      Pipeline per event.

      Founding Revenue Marketing Lead per event

    First result
    2 weeks
    Cost
    Staff time only
    Owner
    Founding Revenue Operations Lead

    Smallest test

    Run the automation for the next event.

    Why it works

    Automate list import, enrichment in Clay, scoring and rep assignment in HubSpot, and send attendees a follow-up within 48 hours with their workshop repo, a Starter link and the next step for their company.

    Speed and relevance decide whether event interest becomes pipeline.

    Risks and how to handle them

    • Bad lists

      Clean and dedupe before routing.

    Evidence

    • Your websiteEvents page: hands-on workshops at AI Engineer World's Fair (build an MCP server and lock it down with Keycard) and Black Hat USA ("Bring a laptop with Node and npm"), an Agent Baseline demo night at AI Engineer New York, AAuth nights in San Francisco hosted with the Insecure Agents podcast, and a panel with Snyk and Docker.Source
    • HiringFounding Revenue Operations Lead (posted October 8, 2026): "We sell top-down to large enterprises and we're launching a developer self-serve motion. Today, an agency manages our GTM systems." Takes over HubSpot and Clay, defines "lead and account scoring, routing, and attribution across sales-led and self-serve paths", automates "product-qualified lead alerts" and "event lead processing from list to rep assignment ahead of our next major events", and makes sure "trial signups and usage signals reach the right rep with context".Source
    • HiringFounding Revenue Marketing Lead (posted October 8, 2026): "turn marketing into predictable pipeline for Sales" through events, demand gen, account-based marketing (targeted ads, direct mail, executive dinners) and "nurture tracks for enterprise prospects and self-serve users", with "follow-up that gets leads to reps within days"; works with the Head of Marketing. A Software Engineer, SDK was also posted on October 23 to build SDKs "across languages (Go, Python, TypeScript, Rust)".Source
  • 07

    Keycard's sign-up form already lists "AI agent" as a way people hear about it, and its buyers build with coding agents daily. Those agents answer "how do I give my agent scoped credentials?" from public docs and pages.

    Medium impactLow effort First result in 2 weeks Head of Marketing

    First action

    Count sign-ups that chose "AI agent" as their source.

    How to measure it

    "AI agent" sourced sign-ups

    Now
    –
    Target
    2x
    By
    2 quarters

    Steps

    1. 1

      Baseline

      Source counts.

      Founding Revenue Operations Lead 1 day

    2. 2

      llms.txt

      Docs and site.

      DevRel 2 days

    3. 3

      Markdown docs

      Plain versions.

      DevRel 1 week

    4. 4

      Answer pages

      Top ten questions.

      DevRel 1 month

    5. 5

      Check assistants

      Monthly test prompts.

      Head of Marketing monthly

    6. 6

      Report

      Source trend.

      Founding Revenue Operations Lead monthly

    First result
    2 weeks
    Cost
    Staff time only
    Owner
    Head of Marketing

    Smallest test

    Publish llms.txt this week.

    Why it works

    Publish an llms.txt and clean Markdown versions of the docs, write direct-answer pages for the top questions (token exchange for agents, protecting an MCP server, revoking agent access), and check monthly what assistants recommend.

    Developers increasingly ask their agent first; the source the agent trusts gets the evaluation.

    Risks and how to handle them

    • Hard to measure

      Combine source answers with referrer data.

    Evidence

    • Your websitePricing page: Starter "Free", "5,000 transactions/mo (hard cap)", "Unlimited users, agents & apps", "Community support"; Team "$500 /mo" with 100,000 transactions and "$1 per 1,000 additional"; Enterprise on annual commitment. Every plan's "Get started" button opens the same early-access form: "Sign up with your work email to request a Keycard account", ending "We'll be in touch". The form asks "How did you hear about us?" with options including "Podcast", "Event" and "AI agent". The site links no community forum or chat for Starter's community support (footer: X, GitHub, LinkedIn, Insecure Agents).Source
    • Your docsDocs navigation: guides to protect MCP servers and APIs, connect coding agents, build a Slack agent, run on Render and Fly.io "without secrets", connect Okta, Entra ID, Ping Identity and Auth0, write, test and roll back policies, plus use cases such as "Same agent, different user" and "Revoke once, access stops"; a docs assistant, "Ask Kai".Source
    • Your websiteHomepage: "Runtime auth for agents"; header buttons "Request access" and "Book a demo"; one customer quote, from Chime ("We had agents running against production systems in days"); SOC 2 Type II and a trust center; the Agent Baseline white paper "developed in partnership with Docker and Snyk"; quotes from Insecure Agents podcast guests at Cursor, Vercel, Mastercard and J.P. Morgan; SDKs for Python, TypeScript, Ruby and Go; federates Okta, Entra, Google and Auth0; a "Talk to our agent" link in the footer.Source

This quarter 9

The core plays for the next three months.

  • 08

    Field engineers run each evaluation by hand today, and the PM is asked to turn that into "eval guides, reference architectures, in-product guidance". Keycard already has the Agent Baseline white paper with Docker and Snyk and a SOC 2 Type II trust center.

    High impactMedium effort First result in 1 quarter Field engineer

    First action

    Write the success criteria from the last five won evaluations.

    How to measure it

    Self-guided evaluation share

    Now
    –
    Target
    50%
    By
    2 quarters

    Steps

    1. 1

      Study wins

      Last five evaluations.

      Field engineer 1 week

    2. 2

      Write the plan

      14-day plan.

      Product Manager, Growth & Developer Experience 2 weeks

    3. 3

      Architectures

      Two diagrams with code.

      Field engineer 2 weeks

    4. 4

      Review pack

      Agent Baseline and trust center.

      Staff Product Marketing Manager 2 weeks

    5. 5

      Checkpoints

      When a human joins.

      Field engineer 1 week

    6. 6

      Measure

      Time to decision.

      Product Manager, Growth & Developer Experience monthly

    First result
    1 quarter
    Cost
    Staff time only
    Owner
    Field engineer

    Smallest test

    Give the plan to the next two evaluators.

    Why it works

    Package a 14-day evaluation plan (milestones, test cases, success criteria), two reference architectures (gateway and in-app SDK), and a security-review pack built on Agent Baseline and the trust center. Offer a field engineer only at defined checkpoints.

    A clear plan shortens evaluations and lets one field engineer support many more.

    Risks and how to handle them

    • Complex deals still need people

      Keep checkpoints for large accounts.

    Evidence

    • HiringProduct Manager, Growth & Developer Experience (remote US, posted August 31, 2026): "Every customer starts with an evaluation, and today a field engineer runs each one by hand: setup, first credential exchange, integration questions. Your job is to turn that walkthrough into product, so an evaluation can run well without us in the room." The PM will "Drive time-to-first-credential-exchange down", "instrument the journey so we can see where evaluators stall", and ship fixes that are often "a better quickstart, a template repo, an error message that says what to do next".Source
    • Your websiteHomepage: "Runtime auth for agents"; header buttons "Request access" and "Book a demo"; one customer quote, from Chime ("We had agents running against production systems in days"); SOC 2 Type II and a trust center; the Agent Baseline white paper "developed in partnership with Docker and Snyk"; quotes from Insecure Agents podcast guests at Cursor, Vercel, Mastercard and J.P. Morgan; SDKs for Python, TypeScript, Ruby and Go; federates Okta, Entra, Google and Auth0; a "Talk to our agent" link in the footer.Source
    • HiringStaff Product Marketing Manager (posted October 8, 2026): "help Keycard win enterprise security buyers" (security architects, CISOs, platform leaders); run "win-loss analysis, track competitors across the agent security landscape"; build competitor comparisons and objection handling, "a launch-tiering framework and playbook", analyst briefings, and in the first 90 days "Publish your first customer case study". Bonus experience: "turning developer usage into enterprise deals".Source
  • 09

    Keycard ships often (coding agents, multi-agent apps, Okta cross-app access, Smallstep, two acquisitions), and the PMM is asked for a launch-tiering framework. Without tiers every release gets the same blog post.

    Medium impactMedium effort First result in 1 quarter Staff Product Marketing Manag…

    First action

    Tier the last eight releases after the fact.

    How to measure it

    Pipeline per tier-1 launch

    Now
    –
    Target
    Measured
    By
    2 quarters

    Steps

    1. 1

      Define tiers

      Three tiers.

      Staff Product Marketing Manager 1 week

    2. 2

      Calendar

      Quarterly plan.

      Staff Product Marketing Manager 1 week

    3. 3

      Templates

      Per tier.

      Staff Product Marketing Manager 2 weeks

    4. 4

      Run a launch

      Next tier-1.

      Staff Product Marketing Manager 1 month

    5. 5

      Measure

      Sign-ups and pipeline.

      Founding Revenue Operations Lead per launch

    6. 6

      Improve

      After each launch.

      Staff Product Marketing Manager quarterly

    First result
    1 quarter
    Cost
    Staff time only
    Owner
    Staff Product Marketing Manager

    Smallest test

    Tier the next release.

    Why it works

    Define three tiers (major launch with press and event, feature launch with docs and email, update in the changelog), plan a quarterly calendar, and measure each launch by sign-ups and pipeline.

    Tiers focus effort on the launches that move buyers.

    Risks and how to handle them

    • Over-process

      Keep tier 3 to a changelog line.

    Evidence

    • HiringStaff Product Marketing Manager (posted October 8, 2026): "help Keycard win enterprise security buyers" (security architects, CISOs, platform leaders); run "win-loss analysis, track competitors across the agent security landscape"; build competitor comparisons and objection handling, "a launch-tiering framework and playbook", analyst briefings, and in the first 90 days "Publish your first customer case study". Bonus experience: "turning developer usage into enterprise deals".Source
    • Your websiteSitemap (32 URLs, checked October 23, 2026): home, about, blog, careers, contact, events, Keycard for Code, brand, pricing and legal pages, plus 21 blog posts (launches for coding agents and multi-agent apps, Okta cross-app access support, the Smallstep device identity partnership, the Anchor and Runebook acquisitions, two "I built a ... agent" tutorials). No customer or case-study pages, no comparison pages and no use-case or industry pages.Source
    • News coverageKeycard came out of stealth in October 2025 with $38 million: an $8 million seed co-led by a16z and Boldstart Ventures and a $30 million Series A led by Acrew Capital. In January 2026 CRN named it one of 10 cloud computing startups to watch.Source
  • 10

    The docs already describe distinct jobs (protect an MCP server, govern a coding agent, build a Slack agent that acts per user, run workloads without secrets), but the marketing site has only one use-case page, Keycard for Code.

    High impactMedium effort First result in 1 quarter Staff Product Marketing Manag…

    First action

    Pick the use case with the most evaluations.

    How to measure it

    Use-case page sign-ups

    Now
    –
    Target
    25% of sign-ups
    By
    2 quarters

    Steps

    1. 1

      Pick four

      From evaluations.

      Staff Product Marketing Manager 1 week

    2. 2

      Write pages

      Problem, diagram, template.

      Staff Product Marketing Manager 1 month

    3. 3

      Add demos

      Short videos.

      DevRel 1 month

    4. 4

      Publish

      With start-free.

      Head of Marketing 1 week

    5. 5

      Point traffic

      Ads and follow-ups.

      Founding Revenue Marketing Lead ongoing

    6. 6

      Measure

      Sign-ups by page.

      Founding Revenue Operations Lead monthly

    First result
    1 quarter
    Cost
    Staff time only
    Owner
    Staff Product Marketing Manager

    Smallest test

    Publish the MCP server page first.

    Why it works

    Publish a page for each job with the problem, a diagram, a template, a customer or demo, and a start-free button. Point ads and event follow-ups at the matching page.

    Buyers search by problem; a page per job converts better than a general homepage.

    Risks and how to handle them

    • Thin pages

      Each needs a working template.

    Evidence

    • Your websiteSitemap (32 URLs, checked October 23, 2026): home, about, blog, careers, contact, events, Keycard for Code, brand, pricing and legal pages, plus 21 blog posts (launches for coding agents and multi-agent apps, Okta cross-app access support, the Smallstep device identity partnership, the Anchor and Runebook acquisitions, two "I built a ... agent" tutorials). No customer or case-study pages, no comparison pages and no use-case or industry pages.Source
    • Your websiteKeycard for Code page: a native integration for popular coding agents that "resolves identity first so tools, skills, memories, and credentials can be scoped to the work being done", replacing "Over-scoped environment variables copied across machines, tools, and sessions".Source
    • Your docsDocs navigation: guides to protect MCP servers and APIs, connect coding agents, build a Slack agent, run on Render and Fly.io "without secrets", connect Okta, Entra ID, Ping Identity and Auth0, write, test and roll back policies, plus use cases such as "Same agent, different user" and "Revoke once, access stops"; a docs assistant, "Ask Kai".Source
  • 11

    The podcast already hosts security leaders from large companies, and the Revenue Marketing Lead is asked to run account-based programs including executive dinners. Guests and listeners are the security architects Keycard sells to.

    High impactMedium effort First result in 1 quarter Founding Revenue Marketing Le…

    First action

    Pick a city and a guest host.

    How to measure it

    Opportunities per dinner

    Now
    –
    Target
    3
    By
    2 quarters

    Steps

    1. 1

      Pick cities

      Where target accounts are.

      Founding Revenue Marketing Lead 1 week

    2. 2

      Book hosts

      Past guests.

      CEO 2 weeks

    3. 3

      Invite

      Target-account leaders.

      Founding Revenue Marketing Lead 3 weeks

    4. 4

      Host

      Dinner and recording.

      CEO 1 day

    5. 5

      Follow up

      Evaluation kit.

      Founding Revenue Marketing Lead 1 week

    6. 6

      Measure

      Opportunities.

      Founding Revenue Operations Lead per dinner

    First result
    1 quarter
    Cost
    Dinner costs
    Owner
    Founding Revenue Marketing Lead

    Smallest test

    Host one dinner.

    Why it works

    Run small dinners in two cities each quarter around a podcast recording, invite target-account security leaders with a past guest as host, and follow up with the evaluation kit.

    Peer conversations earn trust with CISOs that ads cannot.

    Risks and how to handle them

    • Low attendance

      Keep it to 12 seats.

    Evidence

    • Your websiteHomepage: "Runtime auth for agents"; header buttons "Request access" and "Book a demo"; one customer quote, from Chime ("We had agents running against production systems in days"); SOC 2 Type II and a trust center; the Agent Baseline white paper "developed in partnership with Docker and Snyk"; quotes from Insecure Agents podcast guests at Cursor, Vercel, Mastercard and J.P. Morgan; SDKs for Python, TypeScript, Ruby and Go; federates Okta, Entra, Google and Auth0; a "Talk to our agent" link in the footer.Source
    • HiringFounding Revenue Marketing Lead (posted October 8, 2026): "turn marketing into predictable pipeline for Sales" through events, demand gen, account-based marketing (targeted ads, direct mail, executive dinners) and "nurture tracks for enterprise prospects and self-serve users", with "follow-up that gets leads to reps within days"; works with the Head of Marketing. A Software Engineer, SDK was also posted on October 23 to build SDKs "across languages (Go, Python, TypeScript, Rust)".Source
    • Your websiteEvents page: hands-on workshops at AI Engineer World's Fair (build an MCP server and lock it down with Keycard) and Black Hat USA ("Bring a laptop with Node and npm"), an Agent Baseline demo night at AI Engineer New York, AAuth nights in San Francisco hosted with the Insecure Agents podcast, and a panel with Snyk and Docker.Source
  • 12

    Starter has a 5,000-transaction hard cap, and Team adds SSO and longer telemetry. The RevOps post asks for PQL alerts and usage signals reaching reps. A team that hits the cap or asks for SSO is ready to talk.

    High impactMedium effort First result in 1 quarter Founding Revenue Operations L…

    First action

    Define three PQL signals.

    How to measure it

    PQL-to-paid rate

    Now
    –
    Target
    20%
    By
    2 quarters

    Steps

    1. 1

      Define signals

      Three signals.

      Founding Revenue Operations Lead 1 week

    2. 2

      Pipe data

      Usage to HubSpot.

      Software Engineer, Data 2 weeks

    3. 3

      Alerts

      At 80% of the cap.

      Founding Revenue Operations Lead 1 week

    4. 4

      In-console prompt

      Upgrade path.

      Product Manager, Growth & Developer Experience 2 weeks

    5. 5

      Rep playbook

      What to say.

      Staff Product Marketing Manager 1 week

    6. 6

      Measure

      Conversion.

      Founding Revenue Operations Lead monthly

    First result
    1 quarter
    Cost
    Engineering time
    Owner
    Founding Revenue Operations Lead

    Smallest test

    Alert reps on the top ten Starter orgs by hand.

    Why it works

    Score Starter orgs on transactions, users, resources and SSO attempts; alert the rep at 80% of the cap with context; and show an in-console upgrade prompt.

    Usage tells sales who to call and when.

    Risks and how to handle them

    • Spammy outreach

      Contact only on strong signals.

    Evidence

    • Your websitePricing page: Starter "Free", "5,000 transactions/mo (hard cap)", "Unlimited users, agents & apps", "Community support"; Team "$500 /mo" with 100,000 transactions and "$1 per 1,000 additional"; Enterprise on annual commitment. Every plan's "Get started" button opens the same early-access form: "Sign up with your work email to request a Keycard account", ending "We'll be in touch". The form asks "How did you hear about us?" with options including "Podcast", "Event" and "AI agent". The site links no community forum or chat for Starter's community support (footer: X, GitHub, LinkedIn, Insecure Agents).Source
    • HiringFounding Revenue Operations Lead (posted October 8, 2026): "We sell top-down to large enterprises and we're launching a developer self-serve motion. Today, an agency manages our GTM systems." Takes over HubSpot and Clay, defines "lead and account scoring, routing, and attribution across sales-led and self-serve paths", automates "product-qualified lead alerts" and "event lead processing from list to rep assignment ahead of our next major events", and makes sure "trial signups and usage signals reach the right rep with context".Source
    • HiringProduct Manager, Growth & Developer Experience (remote US, posted August 31, 2026): "Every customer starts with an evaluation, and today a field engineer runs each one by hand: setup, first credential exchange, integration questions. Your job is to turn that walkthrough into product, so an evaluation can run well without us in the room." The PM will "Drive time-to-first-credential-exchange down", "instrument the journey so we can see where evaluators stall", and ship fixes that are often "a better quickstart, a template repo, an error message that says what to do next".Source
  • 13

    The homepage quotes podcast guests from Cursor, Vercel, Mastercard and J.P. Morgan and links out to the episodes, and the sign-up form counts "Podcast" as a source. The sitemap has no episode pages.

    Medium impactLow effort First result in 1 month Head of Marketing

    First action

    Publish the five most-listened episodes.

    How to measure it

    Episode-page sign-ups

    Now
    –
    Target
    50 a quarter
    By
    2 quarters

    Steps

    1. 1

      Pick episodes

      Top five.

      Head of Marketing 1 day

    2. 2

      Transcribe

      Clean transcripts.

      Head of Marketing 1 week

    3. 3

      Write takeaways

      Per episode.

      Staff Product Marketing Manager 1 week

    4. 4

      Publish

      Episode pages.

      Head of Marketing 1 week

    5. 5

      Link feed

      Show notes.

      Head of Marketing 1 day

    6. 6

      Measure

      Traffic and sign-ups.

      Founding Revenue Operations Lead monthly

    First result
    1 month
    Cost
    Staff time only
    Owner
    Head of Marketing

    Smallest test

    Publish one episode page.

    Why it works

    Publish an episode page with a transcript, key takeaways and a related template or use case for each episode; link from the podcast feed.

    Transcripts rank for the exact questions security leaders discuss, and the traffic lands on Keycard's own site.

    Risks and how to handle them

    • Guest approval

      Share transcripts before publishing.

    Evidence

    • Your websiteHomepage: "Runtime auth for agents"; header buttons "Request access" and "Book a demo"; one customer quote, from Chime ("We had agents running against production systems in days"); SOC 2 Type II and a trust center; the Agent Baseline white paper "developed in partnership with Docker and Snyk"; quotes from Insecure Agents podcast guests at Cursor, Vercel, Mastercard and J.P. Morgan; SDKs for Python, TypeScript, Ruby and Go; federates Okta, Entra, Google and Auth0; a "Talk to our agent" link in the footer.Source
    • Your websiteSitemap (32 URLs, checked October 23, 2026): home, about, blog, careers, contact, events, Keycard for Code, brand, pricing and legal pages, plus 21 blog posts (launches for coding agents and multi-agent apps, Okta cross-app access support, the Smallstep device identity partnership, the Anchor and Runebook acquisitions, two "I built a ... agent" tutorials). No customer or case-study pages, no comparison pages and no use-case or industry pages.Source
    • Your websitePricing page: Starter "Free", "5,000 transactions/mo (hard cap)", "Unlimited users, agents & apps", "Community support"; Team "$500 /mo" with 100,000 transactions and "$1 per 1,000 additional"; Enterprise on annual commitment. Every plan's "Get started" button opens the same early-access form: "Sign up with your work email to request a Keycard account", ending "We'll be in touch". The form asks "How did you hear about us?" with options including "Podcast", "Event" and "AI agent". The site links no community forum or chat for Starter's community support (footer: X, GitHub, LinkedIn, Insecure Agents).Source
  • 14

    The Revenue Marketing Lead owns nurture tracks for "self-serve users", and the PM is instrumenting the evaluation journey. A generic newsletter cannot help someone stuck on connecting their identity provider.

    Medium impactLow effort First result in 1 month Founding Revenue Marketing Le…

    First action

    Write the track for "no credential exchange".

    How to measure it

    Resume rate

    Now
    –
    Target
    30%
    By
    1 quarter

    Steps

    1. 1

      Map stalls

      Four points.

      Product Manager, Growth & Developer Experience 1 week

    2. 2

      Write tracks

      Two emails each.

      Founding Revenue Marketing Lead 2 weeks

    3. 3

      Build in HubSpot

      Triggered by milestones.

      Founding Revenue Operations Lead 1 week

    4. 4

      Offer help

      Call booking.

      Field engineer ongoing

    5. 5

      Test

      Copy and timing.

      Founding Revenue Marketing Lead monthly

    6. 6

      Measure

      Resume rate.

      Founding Revenue Operations Lead monthly

    First result
    1 month
    Cost
    Staff time only
    Owner
    Founding Revenue Marketing Lead

    Smallest test

    Email ten stalled evaluators by hand.

    Why it works

    Build one short track per stall point (no resource added, no application, no credential exchange, no policy), each with the exact doc, template and an offer of a 15-minute call.

    Specific help at the stuck moment beats general content.

    Risks and how to handle them

    • Depends on events

      Ship after the funnel events exist.

    Builds on

    Evidence

    • HiringFounding Revenue Marketing Lead (posted October 8, 2026): "turn marketing into predictable pipeline for Sales" through events, demand gen, account-based marketing (targeted ads, direct mail, executive dinners) and "nurture tracks for enterprise prospects and self-serve users", with "follow-up that gets leads to reps within days"; works with the Head of Marketing. A Software Engineer, SDK was also posted on October 23 to build SDKs "across languages (Go, Python, TypeScript, Rust)".Source
    • HiringProduct Manager, Growth & Developer Experience (remote US, posted August 31, 2026): "Every customer starts with an evaluation, and today a field engineer runs each one by hand: setup, first credential exchange, integration questions. Your job is to turn that walkthrough into product, so an evaluation can run well without us in the room." The PM will "Drive time-to-first-credential-exchange down", "instrument the journey so we can see where evaluators stall", and ship fixes that are often "a better quickstart, a template repo, an error message that says what to do next".Source
    • Your docsDocs navigation: guides to protect MCP servers and APIs, connect coding agents, build a Slack agent, run on Render and Fly.io "without secrets", connect Okta, Entra ID, Ping Identity and Auth0, write, test and roll back policies, plus use cases such as "Same agent, different user" and "Revoke once, access stops"; a docs assistant, "Ask Kai".Source
  • 15

    Keycard supports Okta cross-app access, partnered with Smallstep on device identity, and wrote Agent Baseline with Docker and Snyk. Each partner reaches the same platform and security buyers.

    High impactMedium effort First result in 2 quarters CEO

    First action

    Propose one joint webinar to each partner.

    How to measure it

    Partner-influenced opportunities

    Now
    –
    Target
    20%
    By
    3 quarters

    Steps

    1. 1

      List programs

      Per partner.

      CEO 1 week

    2. 2

      Pitch

      Partner marketing teams.

      Head of Marketing 1 month

    3. 3

      Listings

      Integration directories.

      DevRel 1 month

    4. 4

      Run webinar

      First one.

      Staff Product Marketing Manager 1 quarter

    5. 5

      Joint demos

      At events.

      Founding Revenue Marketing Lead per event

    6. 6

      Measure

      Pipeline.

      Founding Revenue Operations Lead quarterly

    First result
    2 quarters
    Cost
    Staff time only
    Owner
    CEO

    Smallest test

    Run one joint webinar.

    Why it works

    Agree one joint program per partner per half: a webinar, a joint guide, a marketplace or integration-directory listing, and shared event demos. Track sourced and influenced pipeline.

    Partners lend trust and reach in a new category.

    Risks and how to handle them

    • Partner priorities differ

      Start with the partner most active today.

    Evidence

    • Your websiteSitemap (32 URLs, checked October 23, 2026): home, about, blog, careers, contact, events, Keycard for Code, brand, pricing and legal pages, plus 21 blog posts (launches for coding agents and multi-agent apps, Okta cross-app access support, the Smallstep device identity partnership, the Anchor and Runebook acquisitions, two "I built a ... agent" tutorials). No customer or case-study pages, no comparison pages and no use-case or industry pages.Source
    • Your websiteHomepage: "Runtime auth for agents"; header buttons "Request access" and "Book a demo"; one customer quote, from Chime ("We had agents running against production systems in days"); SOC 2 Type II and a trust center; the Agent Baseline white paper "developed in partnership with Docker and Snyk"; quotes from Insecure Agents podcast guests at Cursor, Vercel, Mastercard and J.P. Morgan; SDKs for Python, TypeScript, Ruby and Go; federates Okta, Entra, Google and Auth0; a "Talk to our agent" link in the footer.Source
    • Your websiteEvents page: hands-on workshops at AI Engineer World's Fair (build an MCP server and lock it down with Keycard) and Black Hat USA ("Bring a laptop with Node and npm"), an Agent Baseline demo night at AI Engineer New York, AAuth nights in San Francisco hosted with the Insecure Agents podcast, and a panel with Snyk and Docker.Source
  • 16

    The PMM supports analyst and executive briefings, Keycard was named a CRN startup to watch, and "agent identity is new". Analysts are writing the category definitions buyers will use.

    Medium impactMedium effort First result in 2 quarters Staff Product Marketing Manag…

    First action

    List five analysts covering non-human identity.

    How to measure it

    Report mentions

    Now
    –
    Target
    3
    By
    3 quarters

    Steps

    1. 1

      List analysts

      Five.

      Staff Product Marketing Manager 1 week

    2. 2

      Write briefing

      Category view.

      Staff Product Marketing Manager 2 weeks

    3. 3

      Book

      Briefings.

      Staff Product Marketing Manager 1 month

    4. 4

      Brief

      With the CEO.

      CEO 1 month

    5. 5

      Follow up

      Proof points.

      Staff Product Marketing Manager ongoing

    6. 6

      Track

      Inclusions.

      Staff Product Marketing Manager quarterly

    First result
    2 quarters
    Cost
    Staff time only
    Owner
    Staff Product Marketing Manager

    Smallest test

    Book one briefing.

    Why it works

    Brief the main security analysts twice a year with a clear category view, Agent Baseline and customer proof, and track inclusion in reports.

    Being in the first category reports shapes enterprise shortlists for years.

    Risks and how to handle them

    • Slow payoff

      Treat as a long-term bet.

    Evidence

    • HiringStaff Product Marketing Manager (posted October 8, 2026): "help Keycard win enterprise security buyers" (security architects, CISOs, platform leaders); run "win-loss analysis, track competitors across the agent security landscape"; build competitor comparisons and objection handling, "a launch-tiering framework and playbook", analyst briefings, and in the first 90 days "Publish your first customer case study". Bonus experience: "turning developer usage into enterprise deals".Source
    • News coverageKeycard came out of stealth in October 2025 with $38 million: an $8 million seed co-led by a16z and Boldstart Ventures and a $30 million Series A led by Acrew Capital. In January 2026 CRN named it one of 10 cloud computing startups to watch.Source
    • Your websiteHomepage: "Runtime auth for agents"; header buttons "Request access" and "Book a demo"; one customer quote, from Chime ("We had agents running against production systems in days"); SOC 2 Type II and a trust center; the Agent Baseline white paper "developed in partnership with Docker and Snyk"; quotes from Insecure Agents podcast guests at Cursor, Vercel, Mastercard and J.P. Morgan; SDKs for Python, TypeScript, Ruby and Go; federates Okta, Entra, Google and Auth0; a "Talk to our agent" link in the footer.Source

Big bets 3

Larger plays that can change the trajectory.

  • 17

    Keycard for Code targets "Over-scoped environment variables copied across machines, tools, and sessions". Developers do not know how many standing credentials their agents hold until they look.

    High impactHigh effort First result in 3 quarters Software Engineer

    First action

    Prototype detection for three common agent config files.

    How to measure it

    Scanner-to-sign-up rate

    Now
    –
    Target
    10%
    By
    4 quarters

    Steps

    1. 1

      Scope

      Config files to scan.

      Software Engineer, SDK 2 weeks

    2. 2

      Build CLI

      Local only.

      Software Engineer, SDK 1 quarter

    3. 3

      Security review

      No data leaves.

      CTO 1 month

    4. 4

      Release

      Open source.

      DevRel 1 week

    5. 5

      Promote

      Workshops and podcast.

      Founding Revenue Marketing Lead 1 quarter

    6. 6

      Measure

      Runs and sign-ups.

      Product Manager, Growth & Developer Experience monthly

    First result
    3 quarters
    Cost
    Engineering time
    Owner
    Software Engineer, SDK

    Smallest test

    Release a scanner for one config format.

    Why it works

    Release a CLI that scans a laptop or repo for agent configs and MCP settings holding long-lived keys, reports them locally (no data leaves the machine), and shows how to replace each with Keycard. Promote it at workshops and in the podcast.

    A free tool that shows the problem on the user's own machine is the strongest top of funnel a security product can have.

    Risks and how to handle them

    • Seen as scareware

      Keep it local, quiet and useful.

    Evidence

    • Your websiteKeycard for Code page: a native integration for popular coding agents that "resolves identity first so tools, skills, memories, and credentials can be scoped to the work being done", replacing "Over-scoped environment variables copied across machines, tools, and sessions".Source
    • GitHubgithub.com/keycardai: about 20 public repositories pushed in October 2026, including the TypeScript, Python, Go and Ruby SDKs, "templates" ("Keycard templates for building secure agents"), getting-started-mcp, a Terraform provider, a GitHub Actions auth action, a Slack agent tutorial and an egress proxy. The most-starred repository has 4 stars; most have 0 or 1.Source
    • Your websiteHomepage: "Runtime auth for agents"; header buttons "Request access" and "Book a demo"; one customer quote, from Chime ("We had agents running against production systems in days"); SOC 2 Type II and a trust center; the Agent Baseline white paper "developed in partnership with Docker and Snyk"; quotes from Insecure Agents podcast guests at Cursor, Vercel, Mastercard and J.P. Morgan; SDKs for Python, TypeScript, Ruby and Go; federates Okta, Entra, Google and Auth0; a "Talk to our agent" link in the footer.Source
  • 18

    Agent Baseline, written with Docker and Snyk, defines six outcomes for deploying enterprise agents and is meant to be brought "to your next security review". A PDF is read once; a scored assessment is shared.

    High impactMedium effort First result in 3 quarters Staff Product Marketing Manag…

    First action

    Write 20 questions mapped to the six outcomes.

    How to measure it

    Completed assessments

    Now
    –
    Target
    1,000
    By
    4 quarters

    Steps

    1. 1

      Write questions

      Mapped to outcomes.

      Staff Product Marketing Manager 2 weeks

    2. 2

      Partner review

      Docker and Snyk.

      Staff Product Marketing Manager 1 month

    3. 3

      Build tool

      Score and gaps.

      Software Engineer, Data 1 month

    4. 4

      Launch

      With partners.

      Head of Marketing 1 month

    5. 5

      Follow up

      Gap-based nurture.

      Founding Revenue Marketing Lead ongoing

    6. 6

      Publish

      Yearly report.

      Staff Product Marketing Manager yearly

    First result
    3 quarters
    Cost
    Staff time only
    Owner
    Staff Product Marketing Manager

    Smallest test

    Run the assessment in a spreadsheet with five prospects.

    Why it works

    Build a short online assessment against the six outcomes, give each team a score and a gap list, and offer a Keycard evaluation for the gaps it closes. Publish aggregate results yearly.

    Assessments create leads with a stated problem and give the category a shared yardstick.

    Risks and how to handle them

    • Partners disagree

      Agree scoring before launch.

    Evidence

    • Your websiteHomepage: "Runtime auth for agents"; header buttons "Request access" and "Book a demo"; one customer quote, from Chime ("We had agents running against production systems in days"); SOC 2 Type II and a trust center; the Agent Baseline white paper "developed in partnership with Docker and Snyk"; quotes from Insecure Agents podcast guests at Cursor, Vercel, Mastercard and J.P. Morgan; SDKs for Python, TypeScript, Ruby and Go; federates Okta, Entra, Google and Auth0; a "Talk to our agent" link in the footer.Source
    • HiringStaff Product Marketing Manager (posted October 8, 2026): "help Keycard win enterprise security buyers" (security architects, CISOs, platform leaders); run "win-loss analysis, track competitors across the agent security landscape"; build competitor comparisons and objection handling, "a launch-tiering framework and playbook", analyst briefings, and in the first 90 days "Publish your first customer case study". Bonus experience: "turning developer usage into enterprise deals".Source
    • Your websiteEvents page: hands-on workshops at AI Engineer World's Fair (build an MCP server and lock it down with Keycard) and Black Hat USA ("Bring a laptop with Node and npm"), an Agent Baseline demo night at AI Engineer New York, AAuth nights in San Francisco hosted with the Insecure Agents podcast, and a panel with Snyk and Docker.Source
  • 19

    Enterprise is sold on annual commitment with dedicated, BYOC or on-prem options, and the RevOps lead will set up deal desk processes during the move upmarket. Large buyers often prefer to buy through their cloud marketplace.

    Medium impactMedium effort First result in 3 quarters Founding Revenue Operations L…

    First action

    Ask the last five enterprise buyers how they would prefer to buy.

    How to measure it

    Marketplace deal share

    Now
    –
    Target
    25%
    By
    4 quarters

    Steps

    1. 1

      Ask buyers

      Preference.

      Founding Revenue Operations Lead 2 weeks

    2. 2

      Pick marketplace

      First one.

      CEO 1 week

    3. 3

      List

      Listing and private offers.

      Founding Revenue Operations Lead 1 quarter

    4. 4

      Train reps

      When to offer.

      Staff Product Marketing Manager 2 weeks

    5. 5

      Run deals

      First three.

      Account Executive 2 quarters

    6. 6

      Measure

      Cycle time.

      Founding Revenue Operations Lead quarterly

    First result
    3 quarters
    Cost
    Marketplace fees
    Owner
    Founding Revenue Operations Lead

    Smallest test

    Run one private offer.

    Why it works

    List Keycard on the major cloud marketplaces with private offers, train reps to offer it, and track deal cycle time against direct deals.

    Buying through existing commitments removes procurement steps.

    Risks and how to handle them

    • Fees cut margin

      Use for deals where speed matters.

    Evidence

    • Your websitePricing page: Starter "Free", "5,000 transactions/mo (hard cap)", "Unlimited users, agents & apps", "Community support"; Team "$500 /mo" with 100,000 transactions and "$1 per 1,000 additional"; Enterprise on annual commitment. Every plan's "Get started" button opens the same early-access form: "Sign up with your work email to request a Keycard account", ending "We'll be in touch". The form asks "How did you hear about us?" with options including "Podcast", "Event" and "AI agent". The site links no community forum or chat for Starter's community support (footer: X, GitHub, LinkedIn, Insecure Agents).Source
    • HiringFounding Revenue Operations Lead (posted October 8, 2026): "We sell top-down to large enterprises and we're launching a developer self-serve motion. Today, an agency manages our GTM systems." Takes over HubSpot and Clay, defines "lead and account scoring, routing, and attribution across sales-led and self-serve paths", automates "product-qualified lead alerts" and "event lead processing from list to rep assignment ahead of our next major events", and makes sure "trial signups and usage signals reach the right rep with context".Source
    • HiringStaff Product Marketing Manager (posted October 8, 2026): "help Keycard win enterprise security buyers" (security architects, CISOs, platform leaders); run "win-loss analysis, track competitors across the agent security landscape"; build competitor comparisons and objection handling, "a launch-tiering framework and playbook", analyst briefings, and in the first 90 days "Publish your first customer case study". Bonus experience: "turning developer usage into enterprise deals".Source

Wildcards 3

Unexpected experiments, each with a cheap way to test it.

  • 20

    The footer links to a Keycard agent visitors can talk to. A company selling runtime auth for agents can show its own agent being authorized in real time.

    Medium impactLow effort First result in 1 month DevRel

    First action

    Log one session's credential events and display them.

    How to measure it

    Agent-to-sign-up rate

    Now
    –
    Target
    5%
    By
    2 quarters

    Steps

    1. 1

      Design view

      Session audit panel.

      DevRel 1 week

    2. 2

      Expose events

      Safe, redacted view.

      Software Engineer, Systems 2 weeks

    3. 3

      Link template

      Build your own.

      DevRel 1 day

    4. 4

      Promote

      Homepage and events.

      Head of Marketing 1 week

    5. 5

      Measure

      Sessions to sign-ups.

      Product Manager, Growth & Developer Experience monthly

    6. 6

      Improve

      Monthly.

      DevRel monthly

    First result
    1 month
    Cost
    Engineering time
    Owner
    DevRel

    Smallest test

    Add a screenshot of the agent's audit log to the homepage.

    Why it works

    Show visitors which credentials the agent requested, what policy allowed or denied, and the audit events of their own session, with a link to build the same thing from a template.

    Seeing the product work on the site is more convincing than any diagram.

    Risks and how to handle them

    • Exposing too much

      Show only redacted, session-scoped events.

    Evidence

    • Your websiteHomepage: "Runtime auth for agents"; header buttons "Request access" and "Book a demo"; one customer quote, from Chime ("We had agents running against production systems in days"); SOC 2 Type II and a trust center; the Agent Baseline white paper "developed in partnership with Docker and Snyk"; quotes from Insecure Agents podcast guests at Cursor, Vercel, Mastercard and J.P. Morgan; SDKs for Python, TypeScript, Ruby and Go; federates Okta, Entra, Google and Auth0; a "Talk to our agent" link in the footer.Source
    • Your docsDocs navigation: guides to protect MCP servers and APIs, connect coding agents, build a Slack agent, run on Render and Fly.io "without secrets", connect Okta, Entra ID, Ping Identity and Auth0, write, test and roll back policies, plus use cases such as "Same agent, different user" and "Revoke once, access stops"; a docs assistant, "Ask Kai".Source
    • GitHubgithub.com/keycardai: about 20 public repositories pushed in October 2026, including the TypeScript, Python, Go and Ruby SDKs, "templates" ("Keycard templates for building secure agents"), getting-started-mcp, a Terraform provider, a GitHub Actions auth action, a Slack agent tutorial and an egress proxy. The most-starred repository has 4 stars; most have 0 or 1.Source
  • 21

    Keycard has already written about building an agent to run event raffles. At conferences, a raffle is usually a badge scan with no product contact.

    Medium impactLow effort First result in 1 month Founding Revenue Marketing Le…

    First action

    Reuse the raffle agent at the next event.

    How to measure it

    Entry-to-sign-up rate

    Now
    –
    Target
    25%
    By
    2 quarters

    Steps

    1. 1

      Package agent

      Template repo.

      DevRel 1 week

    2. 2

      Booth flow

      Two-minute entry.

      Founding Revenue Marketing Lead 1 week

    3. 3

      Signage

      QR to repo.

      Head of Marketing 1 week

    4. 4

      Run

      Next event.

      Founding Revenue Marketing Lead per event

    5. 5

      Follow up

      Within 48 hours.

      Founding Revenue Operations Lead per event

    6. 6

      Measure

      Sign-ups.

      Founding Revenue Operations Lead per event

    First result
    1 month
    Cost
    Prizes
    Owner
    Founding Revenue Marketing Lead

    Smallest test

    Run it at one event.

    Why it works

    Run each booth raffle through a small agent that attendees call from their own laptop with a Keycard-issued credential; entering takes two minutes and leaves them with a working template.

    Attendees remember what they built, and every entry is a qualified developer.

    Risks and how to handle them

    • Wi-fi problems

      Offer a phone-friendly fallback.

    Evidence

    • Your websiteSitemap (32 URLs, checked October 23, 2026): home, about, blog, careers, contact, events, Keycard for Code, brand, pricing and legal pages, plus 21 blog posts (launches for coding agents and multi-agent apps, Okta cross-app access support, the Smallstep device identity partnership, the Anchor and Runebook acquisitions, two "I built a ... agent" tutorials). No customer or case-study pages, no comparison pages and no use-case or industry pages.Source
    • Your websiteEvents page: hands-on workshops at AI Engineer World's Fair (build an MCP server and lock it down with Keycard) and Black Hat USA ("Bring a laptop with Node and npm"), an Agent Baseline demo night at AI Engineer New York, AAuth nights in San Francisco hosted with the Insecure Agents podcast, and a panel with Snyk and Docker.Source
    • HiringFounding Revenue Marketing Lead (posted October 8, 2026): "turn marketing into predictable pipeline for Sales" through events, demand gen, account-based marketing (targeted ads, direct mail, executive dinners) and "nurture tracks for enterprise prospects and self-serve users", with "follow-up that gets leads to reps within days"; works with the Head of Marketing. A Software Engineer, SDK was also posted on October 23 to build SDKs "across languages (Go, Python, TypeScript, Rust)".Source
  • 22

    Starter includes "Community support", but keycard.ai links no forum, Slack or Discord, and GitHub repos get few stars. Self-serve users need a place to ask and to see others building.

    Medium impactLow effort First result in 1 month DevRel

    First action

    Pick the platform and seed it with ten common questions.

    How to measure it

    Weekly active members

    Now
    –
    Target
    200
    By
    3 quarters

    Steps

    1. 1

      Pick platform

      Forum or Slack.

      DevRel 1 week

    2. 2

      Seed

      Ten questions.

      DevRel 1 week

    3. 3

      Link

      Console, docs, pricing.

      Product Manager, Growth & Developer Experience 2 days

    4. 4

      Staff

      Daily rotation.

      DevRel daily

    5. 5

      Feature builds

      Monthly.

      DevRel monthly

    6. 6

      Feed docs

      Gaps to the PM.

      Product Manager, Growth & Developer Experience monthly

    First result
    1 month
    Cost
    Staff time only
    Owner
    DevRel

    Smallest test

    Open a channel for the next workshop cohort.

    Why it works

    Open a community space (forum or Slack) linked from the console, docs and pricing; staff it daily with DevRel and engineers; feature community-built templates monthly.

    Community answers scale support for free users and reveal docs gaps for the PM.

    Risks and how to handle them

    • Empty room

      Invite workshop attendees first.

    Evidence

    • Your websitePricing page: Starter "Free", "5,000 transactions/mo (hard cap)", "Unlimited users, agents & apps", "Community support"; Team "$500 /mo" with 100,000 transactions and "$1 per 1,000 additional"; Enterprise on annual commitment. Every plan's "Get started" button opens the same early-access form: "Sign up with your work email to request a Keycard account", ending "We'll be in touch". The form asks "How did you hear about us?" with options including "Podcast", "Event" and "AI agent". The site links no community forum or chat for Starter's community support (footer: X, GitHub, LinkedIn, Insecure Agents).Source
    • GitHubgithub.com/keycardai: about 20 public repositories pushed in October 2026, including the TypeScript, Python, Go and Ruby SDKs, "templates" ("Keycard templates for building secure agents"), getting-started-mcp, a Terraform provider, a GitHub Actions auth action, a Slack agent tutorial and an egress proxy. The most-starred repository has 4 stars; most have 0 or 1.Source
    • HiringProduct Manager, Growth & Developer Experience (remote US, posted August 31, 2026): "Every customer starts with an evaluation, and today a field engineer runs each one by hand: setup, first credential exchange, integration questions. Your job is to turn that walkthrough into product, so an evaluation can run well without us in the room." The PM will "Drive time-to-first-credential-exchange down", "instrument the journey so we can see where evaluators stall", and ship fixes that are often "a better quickstart, a template repo, an error message that says what to do next".Source

03

How Programmable can help

Where an idea in this playbook is something Programmable builds or runs, it's listed here with what we'd do. Everything else is yours to run.

ProgrammableTalk

Fill the events in this plan

These ideas are events. Talk gets registrants for them and tells you who's coming and what their company does before the day.

Plan an event (opens programmable.talk in a new tab)
More example playbooks and reviews
ProgrammableDigital

Show Keycard’s product working

Turn screenshots or the website into an animated product demo in its own brand, for the homepage, launch posts and outreach in this playbook.

Make a demo (opens programmable.digital in a new tab)

More growth playbooks

Other companies we have researched, each read from what they are hiring for.

  • The Gund Company growth strategyManufacturer and fabricator of engineered materials: thermoset composites, thermoplastics, flexible insulation and elastomers, made into…22 growth ideas
  • Karumi growth strategyAn AI demo agent that runs a live demo of a company's real product, not a recording, on the website, on video calls, inside the product…22 growth ideas
  • Royal Change growth strategySydney Cummings Houdyshell's fitness company: thousands of free, ad-free workouts on YouTube, a monthly $8.99 workout and stretch…20 growth ideas
  • LemonLime growth strategyAutomated sales and marketing for small businesses. Every morning it delivers new leads with drafted outreach, branded content and growth…22 growth ideas
  • CareSwift growth strategyAI-assisted EMS narrative software for agencies. Crews enter call details, answer clarifying questions built from the agency's own…23 growth ideas
  • TrophySmack growth strategyMakes custom fantasy football trophies, championship belts, rings, turnover chains, draft boards, loser awards and WallSmack metal art in…21 growth ideas
  • SafeRack growth strategyA US manufacturer of safe access, bulk loading and fall protection equipment: truck, railcar and marine loading platforms and gangways…20 growth ideas
  • Modica growth strategyMakes award-winning cocktail and mocktail mixers created by a professional mixologist: Classic and Cucumber Margarita, Passionfruit…22 growth ideas

See every playbook and website review

Independent research. Programmable wrote this playbook from public sources, as they stood on October 26, 2026, and cites them throughout. Keycard did not commission, review or endorse it, and Programmable is not affiliated with Keycard. Company and product names are trademarks of their owners, used only to identify them. The ideas are suggestions and opinion, not professional, legal or financial advice, and not statements about how Keycard runs its business.

Something wrong, or do you represent Keycard? Email [email protected] and we will correct it or take it down.

30/60/90 roadmap

What to focus on in each month, and the ideas that carry it.

  1. Days 1 to 307 ideas

    Instant Starter access, funnel events, the Chime case study, comparison pages, a template gallery, 48-hour event follow-up and pages written for AI agents.

  2. Days 31 to 6012 ideas

    Evaluation kit, launch tiers, use-case pages, executive dinners, PQLs, podcast pages, step-based nurtures, partner co-marketing and analysts; the live agent demo, raffle demos and a community.

  3. Days 61 to 903 ideas

    Commit to the big bets: an open-source secrets scanner, the Agent Baseline self-assessment and marketplace buying.

The research behind the plan

3 research areas, each summarised in a few lines. Open any of them for the full detail.

Customers

What customers say

Customers value getting agents into production in days. The most cited problem: agents accumulating privilege.

  • Engineers deploying agents without being identity experts
  • Shared service accounts and long-lived keys
  • In their words: “standing credentials”

Market

Competitive landscape

4 competitors compared: Aembit, Descope, Arcade and Astrix.

  • Gap to exploit: Instant self-serve trials of agent authorization
  • Aembit: Workload-first rather than on-behalf-of-user
  • Descope: Broader CIAM focus

Hiring signals

Keycard is adding a developer self-serve motion to its enterprise sales.

  • Product Manager, Growth & Developer Experience: Evaluations run by hand; time to first credential exchange. Ideas 1, 2, 5, 8.
  • Founding Revenue Operations Lead: Self-serve motion, PQLs, event lead automation. Ideas 1, 6, 12.
  • Founding Revenue Marketing Lead: Events, ABM, nurtures for self-serve users. Ideas 6, 11, 14.

From the team that wrote this playbook

All products
  • ProgrammableAgencyWorking AI systems in days, not quarters (opens programmable.agency in a new tab)
  • ProgrammableInfoAn agent that finds, researches and asks firstNew accounts get 100,000 free enrichment credits with code BETA100K (opens programmable.info in a new tab)
  • ProgrammableDigitalScreenshots in, an animated product demo out (opens programmable.digital in a new tab)
  • ProgrammableWatchKnow the day a supplier falls out of compliance (opens programmable.watch in a new tab)
  • ProgrammableTalkGet registrants for your events, and know who’s coming (opens programmable.talk in a new tab)
  • ProgrammableMediaAn AI note taker, interpreter and webinar host for your calls (opens programmable.media in a new tab)
  • ProgrammablePageText what you need, your customer gets a page that works (opens programmable.page in a new tab)
  • ProgrammableCommunityForums with labelled AI members, never empty (opens programmable.community in a new tab)
ProgrammableMarketing
ExamplesLeaderboardsProductsTermsPrivacyCookiesDPALegalContact

© 2026 AI-Powered Growth Intelligence