Contact Us

AI Agent Readiness Assessment

Score your startup across 13 critical dimensions to understand where you stand — and what to fix before deploying agentic AI.

What Good Readiness Looks Like →
Overall Readiness
0.0 / 5.0
0%
0 / 69 scored
Scoring:
1 — None
2 — Aware
3 — Developing
4 — Operational
5 — Optimized
🧹
Data Quality & Hygiene
6 items
—
avg score
▾
Field Completeness Baseline
1 of 6
What We Look For
Measured completeness rate for critical fields (email, phone, title, company, industry, ARR) across Leads, Contacts, Accounts.
Best Practice
Target 90%+ completeness on fields agents will consume; profile and track monthly.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Duplicate Rate & Dedup Process
2 of 6
What We Look For
Measured duplicate rate across Leads, Contacts, Accounts with active deduplication rules or tooling.
Best Practice
Automated dedup tool (Insycle, RingLead) running weekly with defined match keys and master record rules.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Automated Data Cleansing Tool
3 of 6
What We Look For
Dedicated platform running scheduled automations for formatting, standardization, dedup, and enrichment.
Best Practice
Always-on cleansing tool with recurring automations — not one-time manual cleanups.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Field Standardization
4 of 6
What We Look For
Consistent formatting across picklists, phone numbers, addresses, titles, industry, and country codes.
Best Practice
Automated normalization on create + bulk cleanup quarterly; documented format standards.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Stale Data Management
5 of 6
What We Look For
Process to detect and remediate outdated records (bounced emails, job changes, closed companies).
Best Practice
Scheduled scans at 90/180/365 day thresholds with automated archival or re-enrichment.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Data Quality Scorecard
6 of 6
What We Look For
Dashboard tracking completeness, accuracy, duplicate rate, and decay rate across key objects.
Best Practice
Automated scorecard updated weekly with threshold-based alerts when quality dips.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
🔬
Data Enrichment
3 items
—
avg score
▾
Enrichment Provider Strategy
1 of 3
What We Look For
Active third-party enrichment (ZoomInfo, Apollo, Clearbit) filling firmographic, technographic, and contact gaps.
Best Practice
Enrichment on record create + scheduled bulk re-enrichment quarterly; field-level overwrite rules defined.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Enrichment Automation
2 of 3
What We Look For
Enrichment workflows triggered automatically — not manual CSV uploads.
Best Practice
Event-triggered enrichment (new lead → auto-enrich) plus scheduled bulk passes with source-of-truth hierarchy.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Enrichment Quality Monitoring
3 of 3
What We Look For
Tracking fill rates, match rates, and accuracy of enriched data over time.
Best Practice
Monthly review — track fill rate by field, flag declining match rates, compare vendors.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
🔗
Entity Matching & Resolution
4 items
—
avg score
▾
Lead-to-Account Matching
1 of 4
What We Look For
Automated matching of inbound leads to existing accounts via domain, company name, or email pattern.
Best Practice
Real-time matching on creation using domain + fuzzy name; auto-routes to account owner.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Contact & Lead Deduplication
2 of 4
What We Look For
Cross-object dedup identifying the same person across Lead and Contact records.
Best Practice
Automated cross-object resolution using email + name + company with merge audit trail.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Account Hierarchy Mapping
3 of 4
What We Look For
Parent company → subsidiary → division hierarchy defined and maintained.
Best Practice
Defined hierarchy with automated roll-up of revenue, contacts, and activity to parent.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Buying Committee Mapping
4 of 4
What We Look For
Ability to group decision-makers, champions, and blockers on an opportunity.
Best Practice
Contact Roles or custom buying committee object populated on every opp; tracked by role type.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
🏗️
CRM Foundation
7 items
—
avg score
▾
Object Model Cleanliness
1 of 7
What We Look For
Standard + custom objects well-structured; no orphaned objects or conflicting automations.
Best Practice
Documented object model with ERD; annual audit of custom objects.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Field Architecture & Hygiene
2 of 7
What We Look For
No excessive unused fields; required fields enforced; consistent naming conventions.
Best Practice
Field audit — archive unused, enforce naming convention, document every field in data dictionary.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Automation Inventory
3 of 7
What We Look For
All flows, triggers, workflows documented; no conflicting or redundant automations.
Best Practice
Single automation inventory with trigger conditions, execution order, and owner; reviewed quarterly.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Validation & Data Entry Rules
4 of 7
What We Look For
Validation rules enforcing data quality at point of entry (required fields, format checks).
Best Practice
Validation rules on all critical fields with clear error messages; no workarounds.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Permission Sets & Access Control
5 of 7
What We Look For
Role hierarchy and sharing rules intentionally designed — not everyone is admin.
Best Practice
Principle of least privilege; documented RBAC model; quarterly access reviews.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Record Type & Layout Strategy
6 of 7
What We Look For
Record types and page layouts align to business process — not cluttered.
Best Practice
Each record type maps to a distinct process; layouts show only relevant fields per role.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Reporting Foundation
7 of 7
What We Look For
Core reports/dashboards for pipeline, activity, conversion, and forecast — trusted and used.
Best Practice
Standard report library covering pipeline, velocity, conversion, forecast accuracy.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
📧
Marketing Automation Foundation
6 items
—
avg score
▾
MAP-CRM Sync Health
1 of 6
What We Look For
Sync is reliable — no orphaned records, field mapping documented, sync errors monitored.
Best Practice
Documented field mapping, sync direction rules, conflict resolution; error dashboard reviewed weekly.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Lead Scoring Model
2 of 6
What We Look For
Scoring model combining demographic fit + behavioral engagement; syncs to CRM.
Best Practice
Documented model with MQL/SQL thresholds and decay rules; reviewed quarterly against conversion data.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Lifecycle Stage Definitions
3 of 6
What We Look For
Clear, mutually exclusive stages (Subscriber → MQL → SQL → Opp → Customer) with triggers.
Best Practice
Stages documented with entry/exit criteria; transitions tracked for velocity reporting.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Campaign & Source Tracking
4 of 6
What We Look For
UTM framework and campaign naming conventions tracked consistently across channels.
Best Practice
Standardized UTM taxonomy enforced; first-touch and multi-touch attribution enabled.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Form & Data Ingest Quality
5 of 6
What We Look For
Forms enforce standards (required fields, dropdowns over free text, progressive profiling).
Best Practice
All forms use picklists for key fields, progressive profiling, validation on email/phone.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Email Deliverability
6 of 6
What We Look For
Active monitoring of bounce rate, spam complaints, sender reputation; suppression lists maintained.
Best Practice
Monthly list hygiene; hard bounces auto-suppressed; engagement-based sunsetting.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
📋
Data Governance
5 items
—
avg score
▾
Data Dictionary
1 of 5
What We Look For
Centralized, living data dictionary defining every critical field — name, type, source of truth, owner.
Best Practice
Accessible to all teams; updated with every schema change; reviewed quarterly.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Data Ownership (RACI)
2 of 5
What We Look For
Every critical object and field has a named owner responsible for quality.
Best Practice
RACI matrix for data domains; stewards assigned and accountable.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Data Quality SLAs
3 of 5
What We Look For
Defined thresholds for acceptable quality (completeness >90%, dup rate <3%).
Best Practice
Published SLAs with automated alerting when thresholds breach.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Privacy & Consent
4 of 5
What We Look For
GDPR/CCPA compliance — consent tracked, opt-out honored, retention policies enforced.
Best Practice
Consent fields on person records; automated purge workflows; documented retention policy.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Schema Change Management
5 of 5
What We Look For
Process for requesting, approving, and documenting field/object changes before go-live.
Best Practice
Schema change request process — no cowboy field creation; logged in changelog.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
🔌
Integration & API Readiness
6 items
—
avg score
▾
API Inventory & Docs
1 of 6
What We Look For
All active APIs documented — endpoint, auth method, rate limits, data exchanged, owner.
Best Practice
Centralized API registry; every integration has documented owner, error handling, retry logic.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Integration Health Monitoring
2 of 6
What We Look For
Active monitoring of sync status, error rates, and latency across all integrations.
Best Practice
Health dashboard with alerts on failures; SLA for time-to-resolve sync breaks.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Webhook & Event Architecture
3 of 6
What We Look For
Event-driven integrations documented with trigger conditions and consumers.
Best Practice
Event catalog — what fires, what listens, what happens on failure; no undocumented syncs.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
iPaaS / Middleware Layer
4 of 6
What We Look For
Centralized integration platform (Workato, Tray, Zapier, Make) — not scattered connectors.
Best Practice
Single iPaaS as integration hub; all automations version-controlled with error logging.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Credential Management
5 of 6
What We Look For
API keys and OAuth tokens managed securely — not hardcoded or shared informally.
Best Practice
Secrets vault; credential rotation policy; no keys in code or shared docs.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Agent-Ready API Access
6 of 6
What We Look For
Key systems expose APIs that AI agents can consume (read + write) with appropriate scoping.
Best Practice
Every GTM system has API access enabled with scoped permissions ready for agent consumption.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
🗺️
Enterprise Architecture
4 items
—
avg score
▾
System Architecture Diagram
1 of 4
What We Look For
Documented diagram of all systems, integrations, and data flows across the GTM stack.
Best Practice
Living architecture diagram updated quarterly with all connections and data flow direction.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Data Flow Diagram
2 of 4
What We Look For
Visual mapping of how data moves from capture through transformation to reporting.
Best Practice
Each flow shows source, transformation, destination, sync frequency, and error handling.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Object Model / ERD
3 of 4
What We Look For
Entity-relationship diagram showing object hierarchy and cross-system object mapping.
Best Practice
ERD with cardinality and foreign key mappings; updated with every schema change.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Integration Architecture Diagram
4 of 4
What We Look For
Documented view of middleware/iPaaS showing which platform orchestrates each sync.
Best Practice
Centralized diagram with connectors, sync direction, trigger type, and failover routing.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
🧠
MCP & AI Knowledge Base
6 items
—
avg score
▾
Company Wiki / Knowledge Base
1 of 6
What We Look For
Centralized, structured knowledge base with SOPs, playbooks, product info, ICP definitions, and objection handling.
Best Practice
Single source of truth (Notion, Confluence) organized by domain — this is what agents will reference.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
MCP Server Registry
2 of 6
What We Look For
Inventory of all MCP servers — what tools/data each exposes, auth method, which agents consume it.
Best Practice
Centralized registry maintained alongside API docs; updated when new tools or data sources connect.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
MCP Tool & Schema Docs
3 of 6
What We Look For
Every MCP tool has a published spec — input params, output schema, rate limits, error handling.
Best Practice
Tool specs with name, description, params, return schema, and auth scope — treat like API docs.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
MCP-to-Agent Mapping
4 of 6
What We Look For
Documentation of which AI agents connect to which MCP servers and why.
Best Practice
Dependency map showing connections, fallback behavior, and version pinning.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
MCP Access Control
5 of 6
What We Look For
Defined access control for what data MCP servers can read/write, scoped per agent use case.
Best Practice
Least-privilege per tool; audit logging of agent actions; approval workflows for sensitive ops.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Strategic Context for Agents
6 of 6
What We Look For
Key business context (ICP, personas, pricing, competitive positioning) structured for agent consumption.
Best Practice
Strategic context packaged as structured docs or MCP resources — not buried in slides or tribal knowledge.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
🏢
Data Warehouse & Analytics
3 items
—
avg score
▾
Warehouse Exists or Planned
1 of 3
What We Look For
Data warehouse (Snowflake, BigQuery, Redshift) in place or on roadmap.
Best Practice
Warehouse as single source of truth for cross-system reporting; even lightweight (BigQuery + Fivetran) counts.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Reverse ETL to CRM
2 of 3
What We Look For
Ability to push warehouse-computed insights (scores, segments, LTV) back into CRM.
Best Practice
Reverse ETL tool (Census, Hightouch) syncing computed fields back on a schedule.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Cross-System Reporting
3 of 3
What We Look For
Ability to join data across CRM, MAP, product, billing, support in one place.
Best Practice
At minimum: CRM + MAP + billing joinable; product usage data is next priority.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
🎯
AI Use Case Readiness
5 items
—
avg score
▾
Use Case Inventory
1 of 5
What We Look For
Defined list of AI/agent use cases ranked by impact, feasibility, and data readiness.
Best Practice
Prioritized backlog (AI SDR, deal scoring, routing, content gen, CS health) with data dependency mapping.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Pilot Use Case Selected
2 of 5
What We Look For
One high-impact, low-risk use case identified for first agent deployment.
Best Practice
Pick a case with clean data, clear success metric, and limited blast radius (e.g., enrichment + routing).
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Success Metrics Defined
3 of 5
What We Look For
Each use case has measurable KPIs (time saved, conversion lift, cost reduction).
Best Practice
Baseline metric, target metric, and measurement plan defined before deployment.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Human-in-the-Loop Design
4 of 5
What We Look For
Escalation paths and approval workflows for agent actions requiring human review.
Best Practice
Agents autonomous on low-risk; high-risk actions (emails, deal updates) require human approval.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Agent Testing Framework
5 of 5
What We Look For
Process for testing agent outputs before production (eval sets, shadow mode, A/B).
Best Practice
Shadow mode for testing; eval sets with expected outputs; regression testing on prompt changes.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
⚡
Tech Stack AI Capabilities
10 items
—
avg score
▾
Full Stack AI Feature Inventory
1 of 10
What We Look For
Matrix of every tool mapped to its available AI features, license tier, and activation status.
Best Practice
Living AI Feature Map audited quarterly; assign owner per tool to evaluate new releases within 30 days.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
CRM Native AI
2 of 10
What We Look For
Einstein, Copilot, Breeze — predictive scoring, AI summaries, smart capture — licensed, enabled, used?
Best Practice
Audit every AI feature in your CRM license tier before buying third-party duplicates.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
MAP AI Features
3 of 10
What We Look For
Send-time optimization, predictive content, audience lookalikes, subject line scoring — activated?
Best Practice
Turn on send-time optimization and engagement scoring — zero-config wins in existing license.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Sales Engagement AI
4 of 10
What We Look For
Outreach/Salesloft/Apollo AI — sequence optimization, sentiment detection, recommended actions.
Best Practice
Enable AI sequence optimization and reply sentiment detection; surface winning variants automatically.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Conversation Intelligence AI
5 of 10
What We Look For
Gong/Chorus deal risk alerts, coaching, competitive mentions, topic tracking — actively consumed?
Best Practice
Configure AI deal boards and risk scoring; managers review AI-flagged at-risk deals weekly.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Enrichment & Intent AI
6 of 10
What We Look For
ZoomInfo Copilot, 6sense, Bombora — intent signals, buying committee ID, predictive scoring.
Best Practice
Connect intent scores to CRM scoring so prioritization reflects real-time buyer behavior.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
CS Platform AI
7 of 10
What We Look For
Gainsight/ChurnZero/Totango — health scoring, churn prediction, expansion propensity.
Best Practice
AI health scores blending usage, support, NPS, engagement; automated playbooks on score drops.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
BI & Analytics AI
8 of 10
What We Look For
Tableau Ask Data, Looker Gemini, Power BI Copilot — NL query and anomaly detection.
Best Practice
Enable NL query so non-technical stakeholders self-serve without analyst queue.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Support Platform AI
9 of 10
What We Look For
Zendesk AI, Intercom Fin, Freshdesk Freddy — triage, suggested replies, deflection bots.
Best Practice
Activate AI triage and auto-categorization; measure deflection rate against baseline.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Collaboration Tool AI
10 of 10
What We Look For
Asana Intelligence, Notion AI, Slack AI, Teams Copilot — summarization, smart search.
Best Practice
Enable AI meeting summaries and action item extraction to reduce meeting load.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
🔒
Access Control & Security
4 items
—
avg score
▾
Role-Based Access Control
1 of 4
What We Look For
Intentionally designed role hierarchy with principle of least privilege.
Best Practice
Documented RBAC; quarterly reviews; immediate revocation on role change or offboarding.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Agent-Scoped Permissions
2 of 4
What We Look For
AI agents and integrations have scoped, limited permissions — not admin service accounts.
Best Practice
Dedicated service accounts per agent with minimum required permissions; logged and auditable.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
Audit Logging
3 of 4
What We Look For
All data changes, agent actions, and admin operations logged with who/what/when.
Best Practice
Comprehensive audit trail across CRM, MAP, and agent actions; retained per compliance.
1
None
2
Aware
3
Dev
4
Oper
5
Opt
SSO & MFA Enforcement
4 of 4
What We Look For
Single sign-on and multi-factor auth enforced across all GTM tools.
Best Practice
SSO + MFA everywhere; no shared passwords; service credentials in secrets vault.
1
None
2
Aware
3
Dev
4
Oper
5
Opt

Want a deeper audit with a tactical remediation plan?

Talk to JayOh

We engineer growth systems. JayOh builds the RevOps architecture, GTM execution and data infrastructure that turn messy revenue operations into predictable pipeline for B2B SaaS and PE-backed teams.

Book an Intro Call →
Anthropic

Services

Company

Resources

Industries

© 2026 JayOh. All rights reserved.

Privacy PolicyTerms & Conditions

Linkedin Instagram