Strategic AI Adoption: A Crawl–Walk–Run Guide Using the Effort–Value Matrix
Adopting AI across an organization works best when it’s intentional, incremental, and measurable. This guide outlines a practical, up-and-to-the-right approach using the Effort–Value Matrix and the crawl–walk–run framework. You’ll see how to start small, scale what works, and avoid low-value detours—using Hatz AI tools, templates, and governance features.
The Effort–Value Matrix
Low Effort, Low Value (Crawl)
Fast wins that save minutes, teach fundamentals, and build confidence.
Low Effort, High Value (Walk)
Prebuilt, repeatable workflows and agents for high-frequency processes.
High Effort, High Value (Run)
Custom, cross-system automations and agents with organization-wide impact.
Avoid: Low Value, High Effort
If it’s hard to build and doesn’t save time or drive outcomes, do not prioritize it in your adoption journey
Why this matters: Not everyone starts with the same AI fluency. The matrix helps you align teams on a shared path: bottom-left (crawl) → bottom-right (walk) → top-right (run).
Always start with crawl before trying to reinvent the wheel with AI.
Crawl: Low Effort, Low Value (Always Start Here)
Goal: Build usage habit, reduce friction, and establish safe, governed AI usage.
Typical use cases:
Just use the chat and get comfortable with its features, like tools, chat branching, and image generation.
Image generation for quick assets and mockups. Pick from multiple models to draft slides, simple graphics, or visual concepts fast. Useful, fun, and low setup.
Web search and research assistants. Activate AI-powered search to get current answers quickly.
See: AI Tool Selector (choose and activate web search tools)
https://docs.hatz.ai/en/articles/11900262-ai-tool-selector
Communication helpers. Draft emails, summarize threads, and check your calendar availability via Gmail/Outlook integrations.
Industry chatter monitoring. Scan Reddit, X, and Hacker News for trending topics in your space.
How to set up in Hatz
Enable the Tool Selector to turn on Perplexity, Firecrawl, Tavily, Google Search, or Exa as needed.
https://docs.hatz.ai/en/articles/11900262-ai-tool-selectorCompare and pick the right model for the task.
LLM Comparison: https://docs.hatz.ai/en/articles/10716483-llm-comparison
AI Model Selector: https://docs.hatz.ai/en/articles/11501162-ai-model-selectorLearn the basics of secure chat and tool-calling.
Secure Chat + LLM Basics: https://docs.hatz.ai/en/collections/13010419-secure-chat-llm-basics
Inline Tool Calling: https://docs.hatz.ai/en/articles/11639134-inline-tool-calling
Walk: Low Effort, High Value (Scale What Works)
Goal: Templatize wins for repeatable business processes and share them broadly.
Typical use cases:
Department-ready templates (e.g., HR):
Applicant reviews and resume evaluations
Job descriptions
Policy Q&A with your documents
Talent acquisition assistants
These are prebuilt and require only light customization—tweak prompts, output formats, and connect to your knowledge sources.
How to set up in Hatz
Start in the Workshop to discover and customize templates.
The Hatz Workshop: https://docs.hatz.ai/en/articles/12572765-the-hatz-workshop
Workshop FAQ: https://docs.hatz.ai/en/articles/12281902-workshop-faqUse the Hatz Workshop Assistant to build or modify workflows and agents with natural language.
https://docs.hatz.ai/en/articles/12429780-hatz-workshop-assistantDeploy Agents for specific roles and knowledge scopes; share them with teams.
Agents: https://docs.hatz.ai/en/articles/12602274-agentsConnect integrations (e.g., email, calendar, storage, CRM) to ground outputs in your data.
Integrations: https://docs.hatz.ai/en/collections/13497086-integrations
Run: High Effort, High Value (Design for Leverage)
Goal: Build organization-specific systems that automate complex, cross-domain work.
Typical use cases
Multi-source AI agents that read across your internal knowledge bases and systems to triage requests, enforce policies, and complete tasks.
End-to-end workflows: ingest documents, classify, extract structured data, and push results to downstream tools.
Role-based agents for operations, support, compliance, and finance that integrate with ticketing, ERP, and communication tools.
How to execute in Hatz
Architect end-to-end workflows in the Workshop and combine Steps (document processing, validation, exports).
AI Document Processing Steps and Exports in Workflows: https://docs.hatz.ai/en/articles/12332454-ai-document-processing-steps-and-exports-in-workflowsSelect the right LLMs for reasoning-heavy tasks and governance requirements.
Reasoning Models: https://docs.hatz.ai/en/articles/10885790-reasoning-models
LLM Settings: https://docs.hatz.ai/en/articles/11639035-llm-settingsFor MSPs/Admins, leverage AI-In-A-Box, groups, and tenant controls.
AI-In-A-Box: https://docs.hatz.ai/en/collections/13430268-ai-in-a-box
Groups (MSP Admins): https://docs.hatz.ai/en/articles/12657752-groups
What to Avoid: Low Value, High Effort
Indicators:
Months to integrate for a use case that’s used weekly by one person
Heavy prompt engineering with no measurable time savings or quality lift
No clear KPI or adoption target
Action: De-prioritize and replace with a simpler, higher-leverage variant. If it’s valuable, there’s usually a low-effort prototype you can validate first.
Rollout Blueprint - Up and to the right!
Readiness and guardrails
Establish model access and defaults; define where tool-calling is allowed.
LLM Settings: https://docs.hatz.ai/en/articles/11639035-llm-settingsTurn on the Tool Selector and required integrations for initial teams.
AI Tool Selector: https://docs.hatz.ai/en/articles/11900262-ai-tool-selector
Crawl
Train on secure chat, model selection, and web search tools.
Introduce image gen, research assistants, and email/calendar helpers.
Success metric: weekly active users and time saved anecdotes.
Walk
Adopt department templates (HR, Sales, Support, Ops).
Customize prompts and outputs; connect to your documents.
Success metric: task turnaround time, quality consistency, and share rate across teams.
Run
Build cross-system workflows and role-based agents tied to KPIs.
Add evaluations, logging, and publication policies.
Success metric: hours automated per month, SLA adherence, and user satisfaction.
MSPs and Power Users
Join Feature Fridays and use-case workshops for deep dives and patterns.
Explore AI Education for courses and certifications.
AI Education: https://docs.hatz.ai/en/collections/13010932-ai-education
Quick Summary
Think in matrices: combine Effort and Value to prioritize.
Move up and to the right: crawl → walk → run.
Start with quick wins to build habit; scale department templates; then invest in org-wide automations.
Avoid low value, high effort. Validate with low-effort prototypes first.
Use Hatz’s Workshop, Agents, Tool Selector, and Integrations to operationalize AI safely and quickly.