GTM Cost Control SKILL.md

This is the exact file agents read, including the credit header and licence notice. Save it as gtm-cost-control/SKILL.md.

---
name: gtm-cost-control
description: "Control GTM engineering costs: a cost ledger by tool, provider and workflow, credit attribution by table and column, unit costs from enriched contact to meeting and opportunity, a waste checklist, caching and run conditions, budgets with alerts, a target cost per meeting from deal economics and scale, fix or stop decisions. Use whenever Clay credits or data provider costs rise, someone needs cost per meeting or cost per lead for outbound, wants to cut GTM tool spend, set credit budgets or decide which workflows to scale."
license: CC-BY-4.0
metadata:
  author: Afeez Adebayo
  author_url: https://afeezadebayo.com
  author_linkedin: https://www.linkedin.com/in/afeez-adebayo/
  author_x: https://x.com/za_icon
  source: https://afeezadebayo.com/skills-md/marketing/gtm-engineering/gtm-cost-control/
  category: marketing
  subcategory: gtm-engineering
  version: "1.0.0"
  created: "2026-10-10"
  updated: "2026-10-13"
---

# GTM Cost Control

GTM stacks get expensive quietly. Credits disappear on rows that never qualify, several providers are paid for the same field, AI columns run on every row, unused tables refresh on schedules, and nobody connects spend to results. Without unit costs, teams cannot tell whether a workflow is worth scaling or cutting.

This skill controls the cost of GTM engineering: tracking spend by tool, provider, table and workflow; calculating unit costs from cost per enriched contact to cost per meeting and opportunity; finding waste; reducing credits with run conditions, caching and provider choices; setting budgets and alerts; and deciding which workflows to scale, fix or stop.

## Operating principles
1. **Measure cost per result, not per tool.** Cost per qualified meeting matters more than the monthly bill.
2. **Spend on rows that can convert.** Qualify before enriching or writing.
3. **Pay once.** Cache results and avoid re-enriching unchanged records.
4. **Budget by workflow.** Each workflow has a budget, an owner and alerts.
5. **Cut what does not produce pipeline.** Unit economics decide what scales.

## Cost categories
| Category | Examples |
|---|---|
| Platform | Clay or similar plan, seats |
| Data credits | Enrichment lookups, people search, verification |
| AI | Model usage for research and copy columns |
| Providers | Direct contracts with data vendors |
| Sending | Domains, mailboxes, warm-up tools, sequencer seats |
| CRM and integrations | Seats, add-ons, API limits |
| People time | Building and maintaining workflows |

## Unit cost metrics
| Metric | Formula |
|---|---|
| Cost per enriched contact | Enrichment spend / contacts with valid email |
| Cost per qualified account | Total data spend / accounts in tiers A and B |
| Cost per send | Sending and data spend / emails sent |
| Cost per positive reply | Total workflow spend / positive replies |
| Cost per meeting | Total workflow spend / meetings held |
| Cost per opportunity | Total workflow spend / opportunities created |
| Pipeline per pound | Pipeline value / total spend |

## When to use, and when not to
Use it when credit bills rise, when planning budgets, before scaling a workflow, when comparing providers and in monthly GTM reviews.

Don't use it to choose providers on accuracy alone (`enrichment-waterfalls`) or to report pipeline (`gtm-reporting`), though it uses both.

## Inputs
- Invoices and usage reports from each tool and provider.
- Credit usage by table and column, if available.
- Workflow results: rows, sends, replies, meetings, opportunities.
- Team time spent on GTM engineering.

## Workflow

### Step 1. Build the cost ledger
Monthly spend by category, tool and provider; allocate shared costs to workflows by usage.

### Step 2. Attribute credits
Credits per table and per column; flag the top ten consumers.

### Step 3. Calculate unit costs
For each workflow and segment, using the formulas above.

### Step 4. Find waste
Paid columns without run conditions, enrichment before qualification, duplicate providers, re-enrichment of unchanged rows, AI on low tiers, scheduled refreshes on inactive tables, unused seats and mailboxes.

### Step 5. Apply savings
Run conditions, reordered waterfalls, caching (store results and reuse for a set period), lower-cost AI models for simple tasks, smaller refresh schedules, cancelled seats.

### Step 6. Set budgets and alerts
Monthly credit budget per workflow with alerts at 50%, 80% and 100%.

### Step 7. Decide scale, fix or stop
Compare cost per meeting and opportunity with targets based on deal value and win rate.

### Step 8. Review monthly
Unit cost trends, budget use and decisions.

## Target cost per meeting
A simple ceiling: average deal value x win rate from meetings x acceptable acquisition share. Example: £20,000 deals x 10% of meetings becoming wins x 15% acceptable acquisition cost = £300 maximum cost per meeting. Workflows well below the ceiling can scale; those above it need fixing or stopping.

## Waste checklist
- [ ] Paid columns run only on qualified rows.
- [ ] Waterfalls stop on success.
- [ ] Results are cached and not re-bought within the refresh window.
- [ ] AI runs only where it changes the outcome.
- [ ] Inactive tables have schedules turned off.
- [ ] Seats, mailboxes and providers are used.
- [ ] Test runs use small samples.

## Caching rules
| Data | Refresh window |
|---|---|
| Company firmographics | 90 to 180 days |
| Verified work email | 60 to 90 days, or on bounce |
| Job title and role | 60 to 90 days |
| Technographics | 30 to 90 days |
| Signals | Daily or weekly, depending on the signal |
Before any paid lookup, check whether a fresh enough value already exists in the table or CRM.

## Choosing AI models
Use smaller, cheaper models for simple tasks such as classifying titles, cleaning names or short summaries, and larger models only for research or copy that needs judgement. Test the cheaper model on a sample first; keep it if quality holds.

## Monthly review agenda
1. Spend versus budget by workflow.
2. Unit cost trends (per enriched contact, per meeting, per opportunity).
3. Top credit consumers and changes since last month.
4. Savings actions completed and their effect.
5. Scale, fix or stop decisions with owners.

## Using an AI agent
The agent can parse invoices and usage exports into a ledger, attribute credits, calculate unit costs, run the waste checklist against table specs and draft the monthly review. Contract and budget decisions stay with the owner.

Share the unit-cost dashboard with sales leaders each month.

Re-check vendor contracts at least 60 days before renewal dates.

## Edge cases and judgement calls
| Situation | What to do |
|---|---|
| Long sales cycles | Use cost per opportunity and pipeline per pound as leading measures |
| New workflow with no results yet | Set a test budget and a decision date |
| Annual contracts | Measure utilisation; renegotiate or switch at renewal |
| Shared credits across teams | Allocate by usage; give each team a budget |
| Strategic accounts | Accept higher cost per meeting where deal values justify it |
| Price changes from vendors | Re-run benchmarks and unit costs |

## Handoffs to related skills
- Table run conditions: `clay-table-design`
- Provider order: `enrichment-waterfalls`
- AI usage: `ai-copy-columns`
- Pipeline results: `gtm-reporting`
- Audit: `gtm-engineering-audit`

## Worked example
A team's monthly GTM spend had doubled in six months. The ledger showed 60% of credits used by two tables: one enriched every imported company before filtering, the other ran an AI research column on all tiers. Duplicate providers paid for the same emails, and four inactive tables refreshed daily. After adding run conditions, caching enrichment for 90 days, limiting AI to tiers A and B, removing one provider and turning off schedules, credits fell by about 55% with the same number of meetings. Cost per meeting dropped from £140 to £65, well under the £300 ceiling, so the team scaled the best workflow.

## Common mistakes
- Watching the total bill but not cost per meeting.
- Enriching before qualifying.
- Paying several providers for the same data.
- Leaving schedules running on unused tables.
- No budgets or alerts.

## Output format
```
1. Cost ledger by category, tool and workflow
2. Credit attribution (top consumers)
3. Unit costs by workflow and segment
4. Waste found and savings actions
5. Budgets and alerts
6. Scale, fix or stop decisions
```

## Self-check before you answer
- Are costs tied to results through unit costs?
- Are top credit consumers identified?
- Is there a target cost per meeting from deal economics?
- Are budgets and alerts set per workflow?

---
Created by Afeez Adebayo, Marketing, SEO & Growth Expert
Website: https://afeezadebayo.com
LinkedIn: https://www.linkedin.com/in/afeez-adebayo/
X: https://x.com/za_icon
Source: https://afeezadebayo.com/skills-md/marketing/gtm-engineering/gtm-cost-control/

Licence: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). You may copy, share and adapt this skill, including for commercial use, as long as you credit Afeez Adebayo and link to the source URL above. Keep this notice in any copy or adapted version.