Paid ads workflows

An agent monitors creative fatigue, decides which creatives to retire, generates and reviews refresh variants against brand guidelines.

An ad-platform webhook fires when a variant's hourly spend spikes; the workflow checks its recent CTR in BigQuery, logs a fatigue verdict to Airtable.

Each Monday, builds a week-over-week UTM campaign performance scorecard from BigQuery spend and Stripe revenue and writes a structured page into a Notion database.

Turns a single creative brief submitted via webhook into a grid of Replicate-generated ad-image variants, logs each to Airtable, and posts the contact sheet to Slack for review.

When a creative trips the fatigue thresholds, generates three fresh image variants from the original concept and logs them as draft assets in Airtable for the team to review…

Flags spend-burning fatigued creatives, posts a one-click pause-approval request in Slack, and on approval publishes the queued refresh variants to the ad platforms.

Each morning, joins yesterday's ad spend in BigQuery against Salesforce opportunities by campaign.

Generates ad image variants on demand, posts them to a Slack channel with approve/reject buttons.

When a variant is tagged as the A/B test winner in Airtable, the workflow archives the asset to S3 and publishes it across paid-social platforms, then confirms the push in Slack.

Receives daily ad-spend exports via webhook, normalizes and validates the UTM campaign tags against a canonical naming convention.

Scans BigQuery ad-performance data each morning, flags creatives whose click-through rate has dropped past your threshold over a rolling window.

Nightly, refreshes a campaign-tracker Airtable base with reconciled spend, pipeline, and ROAS per UTM campaign so the marketing team has a live editable scoreboard without…

On a schedule, reads cost-per-acquisition by creative from BigQuery, detects sustained CPA inflation that signals fatigue.

An agent reviews each morning's fatigued-creative report, decides whether to refresh, retire, or hold each one based on spend and trend.

Every Monday, aggregates the prior week's spend and won revenue by marketing channel from BigQuery, drafts a narrative summary with an LLM.

Each week pulls fatigued creatives from BigQuery, drafts a plain-English refresh brief naming what is tiring and why.

On a schedule, compares campaigns that incurred ad spend in BigQuery against campaigns present in Salesforce.

Every morning, joins yesterday's ad spend from BigQuery against Stripe revenue by UTM campaign and posts a per-campaign ROAS table to Slack.

Hourly check that flags any UTM campaign accumulating ad spend in BigQuery with zero matching Stripe revenue over a rolling window, and pages the channel to pause it.

When a variant crosses the fatigue threshold, a CEO-style agent investigates the variant's history, drafts an original refresh brief with concrete new-angle ideas.

Watches ad frequency in BigQuery and, when a creative's frequency exceeds the cap while CTR slides, posts a Slack alert to the growth channel and opens a refresh task on a Monday…

Receives real-time performance pings from your ad platform, detects same-day CTR collapse on a creative, posts a threaded Slack alert.

Scans daily ad performance in BigQuery for creatives whose click-through rate has decayed past a threshold.

Cross-references your Airtable creative tracker with BigQuery performance to find creatives that are both old and decaying, marks them stale in Airtable.

On a weekly schedule, pulls active campaign concepts from Airtable, generates fresh image variants.

On a schedule, reads ad performance results from Airtable, identifies top-performing creatives, generates new variations of the winners with Replicate.

A CEO-style agent takes a plain-language campaign goal, drafts distinct creative concepts, generates image variants for each.

On cart abandonment, tags the contact in HubSpot and pushes them into a retargeting ad audience so paid channels can chase the open cart.

On a weekly schedule, pulls active campaign briefs from Airtable, generates fresh ad variants on Replicate.

Listens for live performance events via webhook, evaluates whether a creative has crossed its hard fatigue threshold.

Pulls daily per-creative ad metrics from BigQuery, flags variants whose 3-day CTR has dropped a set percentage below their rolling 14-day baseline.

Each morning, scans yesterday's creative-level ad metrics in BigQuery, flags creatives where frequency has climbed and CTR has decayed past your thresholds.

Every Monday, joins last week's ad spend (by UTM campaign) against closed and open pipeline in your warehouse, computes cost-per-opportunity and ROAS per campaign.

Twice weekly, flags UTM campaigns whose spend has crossed a budget cap while ROAS stays below target.

Runs daily to find ad spend on UTM campaigns that produced zero matching pipeline records, and alerts the marketing-ops channel so broken or mistagged campaigns get caught fast.

Reads the prior week's spend and conversions per UTM campaign from BigQuery and writes computed CAC, ROAS, and a status flag into the Airtable UTM tracking base…

Watches an Airtable 'Creative Brief' table for new rows, generates four image variants per brief with Replicate, writes the rendered URLs back to Airtable.