GTMIQ
DemoResourcesResumeAdmin
← All resources
Download PDF
GTMIQ · One-pager

How I think about GTM systems

Pete Macias. Ten-plus years in B2B sales, now building the systems behind the pipeline: data first, automation second, AI as assistance with a person in control.

Updated Nov 2026

The problem

Most go-to-market stacks don't fail on tools. They fail in the gaps: contact data that's duplicated or stale, lead rules nobody wrote down, handoffs where context gets lost, and reports nobody trusts. Adding AI on top of that makes the mess faster, not better.

My approach, in order

  1. 1 · Data model
    Define stages, fields and the source of truth before building.
  2. 2 · Clean + enrich
    Dedupe, normalize, fill gaps, watch sync health.
  3. 3 · Score + route
    Fit plus signals, clear rules, handoff SLAs.
  4. 4 · AI assists
    Drafts, briefs, classification; a person approves.
  5. 5 · Measure
    Reply, meeting and cost-per-meeting, fed back into 1-4.

What is built, used, and still being learned

StatusWhat
Used in my live setupClay, Apollo, Attio, Make, Supabase, Claude API, Lovable, and an MCP server (11 read / 4 write tools) connected to Claude, ChatGPT, Grok and Gemini.
Used in my sales careerSalesforce and HubSpot as a hands-on user for several years; quoting with Salesforce CPQ as an account executive; MEDDIC-style qualification; running demos, POCs and scoping calls.
LearningSQL and Python (I build with AI coding tools and am learning to read and write the code myself), email deliverability, agent frameworks, evaluation methods.

I label things this way on purpose. Where I'm a user rather than an admin or engineer, the site says so.

Principles

  1. Drafts, not sends
    AI proposes first lines, briefs and labels. A person approves anything customer-facing that carries risk.
  2. Honest numbers
    Metrics come with their formulas and sample sizes. Demo figures are marked as samples, never passed off as results.
  3. Decision support
    Scores and recommendations explain themselves. The person decides.
  4. Small and testable
    One variable at a time, weekly review, keep what works, stop what doesn't.

What you can see on this site

Fifteen-plus demo pages with sample data: CRM data model, data quality, lead routing, signal triggers, content to outreach, meeting handoff, signal center, AI BDR review queue, deal desk, end-to-end workflow, outbound scorecard, pipeline inspection, CPQ quoting, presales kit, AI agents and evals, and deliverability.

What's next

  • Replace sample figures with my own operating metrics as they accumulate.
  • Finish the AI engineering coursework and add a tested eval set to the agents page.
  • Publish a short case study on how the same system supports a second, very different project (The Film Room case study).
Want to talk it through?

I'm open to GTM Engineer, RevOps, AI Enablement and Solutions Engineer roles.

Connect on LinkedInSee the demo
Pete Macias · linkedin.com/in/petemacias · figures are illustrative sample data
GTMIQ is a personal prototype built by Pete Macias.