Skip to main content

Command Palette

Search for a command to run...

The 90-Day AI Agent Marketplace Adoption Playbook

A weekly plan for launching your AI Agent Marketplace, from zero to organizational standard.

Updated
8 min read
The 90-Day AI Agent Marketplace Adoption Playbook
P
Pushp Vashisht is working as a Software Engineer II at Microsoft, Ireland. For more information pay a visit at: pushp.ovh

Audience: Platform engineers, engineering managers


The Adoption Curve

Most internal tools fail not because they're bad, but because nobody knows they exist. The marketplace has an advantage that it makes AI tools better, but you still need a deliberate rollout.

The curve has three distinct phases, each with its own goals and tactics:

Phase 1: Foundation (Weeks 1-4)

Week 1: Setup

Goal: Marketplace repo exists, CLI works, 3 plugins ready.

Day Task Owner
Mon Fork boilerplate, customize org-wide rules (security, coding standards, compliance) Platform engineer
Tue Build Plugin #1: Code Review Agent (your org's review standards) Platform engineer
Wed Build Plugin #2: Test Writer (your org's test patterns and frameworks) Platform engineer
Thu Build Plugin #3: Onboarding Assistant (your org's setup guides, architecture docs) Platform engineer
Fri Set up CI pipeline, CODEOWNERS, PR template. Test end-to-end flow. Platform engineer

Deliverable: Working marketplace with 3 plugins and a CLI.

Week 2: Champions

Goal: Recruit 5-8 early adopters who will seed organic adoption.

Task Details
Identify champions Pick engineers who are already enthusiastic about AI tools, from different teams
Private demo 30-minute walkthrough: install marketplace, install plugin, show it working
Gather feedback What plugins do they want? What rules are missing? What's confusing?
Fix friction Address the top 3 issues champions raise

Champion selection criteria:

  • Already using AI tools daily

  • Respected by peers (their adoption signals credibility)

  • From different teams (cross-pollination)

  • Mix of senior and mid-level (senior for authority, mid-level for volume)

Week 3: Soft Launch

Goal: Available to anyone, announced to champions' teams only.

Task Details
Announce in champions' team channels "Hey, we've been using this marketplace. Check it out."
Office hours 2x 30-minute slots for live help installing and using plugins
Track metrics Install count, questions asked, pain points
First external PR Encourage one champion to contribute a skill or fix

Week 4: Iterate

Goal: Fix everything that broke in Week 3.

Task Details
Retrospective with champions What worked? What didn't?
Fix top 5 issues Documentation gaps, CLI bugs, confusing rules
Add 1-2 more plugins Based on champion feedback
Prepare launch comms Blog post, demo video, FAQ

Phase 1 Checkpoint:

  • 3+ plugins working and documented

  • 5+ engineers actively using the marketplace

  • 1+ external contribution (not from platform team)

  • Top friction points resolved

  • Launch comms ready


Phase 2: Expansion (Weeks 5-8)

Week 5: Org-Wide Launch

Goal: Everyone knows the marketplace exists and how to start.

Channel Action
All-hands / engineering meeting 5-minute demo: problem, solution, marketplace browse, marketplace install
Email / newsletter Announce with link to onboarding guide
Slack/Teams channel Create #ai-marketplace channel for support and showcase
Wiki / intranet Add marketplace to engineering tools page
README template Add marketplace setup to your org's repo template

Week 6: Team-by-Team Onboarding

Goal: Get 3+ teams actively using the marketplace.

Task Details
Team onboarding sessions 15-minute live session per team: install, configure, demonstrate
Team-specific plugins Help each team identify one plugin that solves THEIR problem
Pair programming Platform engineer pairs with team member to install + customize

Week 7: Contribution Drive

Goal: Get engineers outside the platform team contributing.

Tactic Details
"Plugin of the Month" Recognize the best community-contributed plugin
Hackathon hours 2-hour dedicated time for building marketplace plugins
Plugin ideas board Shared list where anyone can propose plugins (vote on priority)
Mentoring Platform engineer reviews plugin PRs with teaching mindset

Week 8: Governance Activation

Goal: The marketplace governs itself.

Task Details
Enable telemetry dashboard Share usage metrics with engineering leadership
First governance review Review and update rules based on 8 weeks of data
Add CODEOWNERS from other teams Distribute ownership beyond platform team
Deprecation test Deprecate the example-plugin to verify the lifecycle works

Phase 2 Checkpoint:

  • 50%+ of engineering teams have at least one plugin installed

  • 10+ plugins in the marketplace

  • 3+ contributors from outside the platform team

  • Telemetry dashboard live and shared with leadership

  • #ai-marketplace channel is active (questions AND answers)


Phase 3: Standard (Weeks 9-12)

Week 9: Integration with Engineering Processes

Goal: Marketplace becomes part of "how we build software."

Integration Action
New repo template Include marketplace init in the org's repo scaffolding script
Onboarding checklist Add "Install marketplace" to new hire engineering setup guide
Code review guidelines Reference marketplace plugins in review standards
Architecture Decision Records Document the marketplace as an organizational standard

Week 10: Advanced Use Cases

Goal: Expand beyond code to other personas.

Use Case Platform Plugin
Incident response in Teams Copilot Studio incident-responder
Sprint status for PMs Copilot Studio project-status-agent
Production deployment agents Azure AI Foundry deploy-validator
Customer support knowledge Copilot Studio support-assistant

Week 11: Self-Service Maturity

Goal: Platform team is no longer the bottleneck.

Task Details
Authoring workshop Train 10+ engineers to write their own plugins
Self-serve review Qualified contributors can approve plugin PRs
Documentation complete Every question in #ai-marketplace has a docs link answer
Plugin quality metrics Dashboard showing plugin health (usage, freshness, errors)

Week 12: Celebrate and Plan

Goal: Report impact and plan next quarter.

Task Details
Impact report Quarterly business review with ROI numbers (see ROI Calculator post)
Retrospective What worked, what to improve, what to build next
Roadmap Next quarter's plugin priorities based on usage data
Recognition Thank champions and top contributors publicly

Phase 3 Checkpoint:

  • Marketplace is in the new repo template

  • Marketplace is in the new hire onboarding checklist

  • 3+ non-developer use cases live (Copilot Studio, AI Foundry)

  • Self-serve contribution pipeline working

  • First quarterly impact report delivered to leadership

  • Next quarter roadmap prioritized by data


The Champions Program

Champions are the single most important factor in marketplace adoption. Here's how to run it:

Selection (5-8 people)

Criterion Why
Already uses AI tools daily They'll see the value immediately
From different teams Cross-pollination drives discovery
Respected by peers Their endorsement carries weight
Vocal in public channels They'll share wins organically
Mix of seniority levels Senior = credibility, mid = volume

Responsibilities

  • Install marketplace in at least 2 repos

  • Give feedback in the first 2 weeks

  • Answer questions in #ai-marketplace channel

  • Submit at least 1 plugin or plugin improvement

  • Demo the marketplace in their team meeting

Incentives

  • Early access to new plugins and features

  • Credit in the marketplace README (Contributors section)

  • Direct channel to platform team for feature requests

  • Recognition at all-hands

  • (Optional) Swag, gift cards, or hackathon time

Anti-Patterns to Avoid

Anti-Pattern Why It Fails Do This Instead
Launch with 0 plugins "There's nothing here" Seed with 3-5 plugins before launch
Mandate adoption Resentment kills contribution Make it so good people choose it
Platform team owns all plugins Bottleneck, no community Own the framework, not the content
No telemetry Can't prove value, funding dies Instrument from day 1
Perfect before launch You'll never launch Launch rough, iterate weekly
Ignore non-developers Misses 60% of the opportunity Plan Copilot Studio agents in Phase 3

This is a bonus post in the AI Agent Marketplace series.

Enterprise AI Agent Marketplace

Part 9 of 10

Most organizations let every team pick their own AI tool. Some use Claude Code, others use GitHub Copilot, others Cursor or Copilot Studio. The result: duplicated workflows, inconsistent governance, and AI capabilities trapped inside individual teams. This series shows you how to fix that with an internal AI Agent Marketplace: a shared catalog of skills, agents, and rules that every team can install into whichever AI platform they already use. Consistency without forced standardization. Inside you'll find a 3-part core walkthrough (why the pattern matters, how to build one, how to integrate it), 5 platform-specific guides (Claude Code, GitHub Copilot, Copilot Studio, Cursor, Azure AI Foundry), and 5 bonus posts covering ROI modeling, a 90-day adoption playbook, 15 ready-to-build plugin recipes, real case studies across company sizes, and an AI maturity model. A production-ready boilerplate repository ships alongside the series so you can fork and customize on day one. Who it's for: platform engineers, engineering managers, and governance teams who want AI adoption to scale without becoming a sprawl of disconnected experiments.

Up next

15 AI Agent Plugin Recipes to Seed Your Marketplace

Ready-to-build plugin specs for every persona in your organization.