---
name: paid-ads-probe-planning
description: Turn a Consumer AI venture hypothesis into a paid-ads demand probe brief with inputs, outputs, budget, UTMs, telemetry, creative matrix, and kill/scale rules.
version: 1.0.0
author: Hermes Agent
---

# Paid Ads Probe Planning

Use this before launching paid traffic for a Consumer AI Factory venture.

## Input

Required:

- Venture/product name
- Target user/persona
- Pain sentence
- First-use promise
- Landing page URL
- Primary conversion event
- Budget and test duration
- Allowed geos/audience constraints
- Available creative assets or creative direction
- Telemetry destination, usually PostHog project/event names

Optional:

- Prior experiment IDs/results
- Competitor examples
- Existing UTM conventions
- Meta/TikTok/Google account IDs
- Constraints around claims, regulated content, or brand safety

## Output

Produce a paid probe brief with:

1. Campaign objective
2. Target audience hypothesis
3. Offer/CTA
4. UTM naming scheme
5. Creative test matrix
6. Landing/conversion instrumentation checklist
7. Daily budget and stop-loss
8. Kill / iterate / scale rules
9. Reporting cadence
10. Experiment ledger fields to update

## Default Consumer AI probe pattern

For early validation, optimize for learning before platform efficiency:

```text
one sharp audience + one sharp pain + 2-4 creatives + one conversion event + clean UTMs
```

Do not run broad spend until first-value activation is credible.

## UTM convention

Use readable, durable names:

```text
utm_source=meta|tiktok|google|reddit
utm_medium=paid_social|paid_search|paid_display
utm_campaign={venture}_{objective}_{creative_family}_v{n}
utm_content={creative_id_or_angle}
utm_term={adset_or_keyword_group}
```

Example:

```text
utm_campaign=alma_leads_video_static_v1
utm_content=static_v1
```

## Creative matrix

For each creative, define:

- Angle
- User pain shown
- Product promise
- CTA
- Asset path/URL
- Platform placement assumptions
- Claim risk

Avoid generic AI SaaS ad slop. Creative should look native to the audience and grounded in the product/user context.

## Kill / iterate / scale rules

Tiny-budget default:

- Let ads exit review and collect enough spend before judging.
- If spend is near zero, diagnose delivery/review/account issues before changing the offer.
- If CTR/link clicks exist but conversion is zero, inspect landing and telemetry before killing the audience.
- If conversion exists but quality is poor, tighten audience/offer and contact leads manually.
- If conversion and replies/usage appear, run the next creative/audience split.

Do not scale from email/waitlist conversions alone. For Consumer AI, require movement toward first value or live user replies.

## Verification

Before handoff:

1. Landing URL loads.
2. UTM links are correct.
3. Conversion event exists and is queryable.
4. Creative assets are available.
5. Budget/stop-loss is explicit.
6. The next platform-specific skill is selected, e.g. `meta-ads-campaign-launch`.