Use case · Performance marketing
Monitors track pacing across Meta and Google Ads, a Relay recommends what to change, and budget changes wait for the growth lead.
An illustrative performance marketing use case for Agent Serve, the AI workspace where teams build, deploy, and manage AI agents: Google Ads reports, Meta ads through Tool connections, pacing monitors, creative fatigue rules and an ad performance Relay.
At a glance
The problem it removes
Spend is checked in each ad platform once a day, if there is time. A campaign that overpaces by lunch is noticed the next morning, and a creative that stopped working keeps spending until someone reads the numbers.
Step by step
Spend, revenue and ROAS from both platforms arrive in one table and one dashboard.
The ad performance Relay reads the alert, checks the campaign and drafts a budget change with its reason.
The budget change waits as an approval and runs exactly once after a yes.
How it fits together
The Google Ads block reads campaign and ad performance, and Meta’s ads server connects under Tool connections.
A pacing dashboard compares spend to plan by campaign, channel and day.
A monitor re-checks daily spend against plan as often as every 5 minutes.
A plain-language rule marks ads whose click-through falls while frequency climbs.
The ad performance Relay drafts the change and waits for approval before it spends.
How it runs
Agent Serve reads the ad platforms and drafts the change, and nothing that spends money runs until a person approves it.
A scheduled workflow reads Google Ads campaign performance into a table. The Google Ads block is read only.
A monitor on daily spend raises a warning when a campaign runs ahead of plan.
The Relay reads campaign insights through Meta’s ads server, connected under Tool connections and on its allowlist.
The Relay classes the change as spend, so it waits for approval unless it fits a delegated limit.
What to measure
Example figures for a team running Meta and Google Ads across a few dozen campaigns.
−60%
Daily reporting time
Both platforms in one table and one dashboard.
−30%
Overspend against plan
Pacing alerts arrive the same morning.
2 days
Earlier fatigue warnings
Tired creatives flagged before results drop far.
0
Unapproved budget changes
Every spend change has a reviewer.
Illustrative example. Figures show how the workflow is measured; they are not customer results.
In depth
Agent Serve puts the pacing dashboard, the Relay, its spending rules and the follow-up question in one workspace.
The Agent Serve pacing dashboard reads one performance table, so spend, revenue and ROAS mean the same thing in every panel.
An Agent Serve Relay classes every call as a read, write, destructive or spend. Reads run; spend waits for approval.
The growth lead sees the campaign, the new budget and the reason the Agent Serve Relay gave, then approves or declines.
Marketers ask Agent Serve Chat about a campaign, read Meta insights through Tool connections, and get a chart back.
Examples
Three growth teams running paid media reviews on Agent Serve, each with a person approving anything that spends.
An Agent Serve monitor watches spend against plan, and a Relay drafts the morning note with the campaigns to adjust.
An Agent Serve rule flags ads whose click-through falls for three days while frequency rises, and lists them for the creative team.
An Agent Serve Relay with delegated limits shifts small amounts between campaigns, and anything above the cap waits for approval.
Examples show what teams can build in Agent Serve. They describe patterns, not named customer deployments.
Before and after
| Work | Before | With Agent Serve |
|---|---|---|
| Daily numbers | BeforeExported from each ad platform into a spreadsheet. | With Agent ServeWritten to one table and read by one dashboard. |
| Overspend | BeforeNoticed in the next morning’s check. | With Agent ServeA monitor raises an alert as pacing crosses plan. |
| Creative fatigue | BeforeSpotted when results have already fallen. | With Agent ServeA rule flags the trend while there is time to swap. |
| Budget changes | BeforeMade in the ad platform with no record of why. | With Agent ServeDrafted with a reason, approved, run once and audited. |
Governance
Agent Serve keeps reads automatic and holds anything that spends money for a named person, unless it fits limits you set.
Security and governanceAgent Serve Relays hold budget changes and new spend for approval unless the call fits inside delegated limits.
An Agent Serve Relay can use only the integrations and Tool connections on its allowlist.
Relays never send email or book meetings without an approval, under any autonomy setting. An approval covers the exact payload the reviewer saw and expires if nobody decides.
Proposals, approvals, declines and executed actions are written to the workspace audit log, which admins can export.
Admin, write and read roles per workspace, plus permission groups that limit which models, blocks and integrations a team can use.
Every workflow run records each block’s inputs, outputs and errors, so a reviewer can trace how a result was reached.
Use your organisation’s own model provider keys at organisation or workspace level.
Further reading
Overspend shows up on the invoice and underspend shows up as a missed month. This dashboard puts Google Ads and Meta spend against one monthly plan, watches CPA drift and alerts the team, while people make the changes.
Growth marketing · 5 min readPerformance teams spend their mornings pulling the same numbers out of Ads Manager. These field notes cover connecting Meta ads to Agent Serve, asking questions in Chat, and letting a Relay send a daily review that a person acts on.
Performance marketing · 6 min readCompetitor research is the job everyone agrees matters and nobody has time for on a Thursday. This playbook sets up a workflow that checks public sources every week and a Relay that sends a short brief on what changed and why it matters.
Brand & growth strategy · 6 min readQuestions
There is no built-in Meta block. You add Meta’s official ads server under Settings → Tool connections, a standard way for AI agents to use another product’s tools. Chat and allowlisted Relays can then read insights and draft campaigns.
No. The Agent Serve Google Ads block is read only: campaigns, ad groups, ad performance and custom queries. Budget, bid and status changes stay in Google Ads.
Only inside delegated limits you set per action, per day and per week. Outside them, an Agent Serve Relay holds every spend change for approval.
No. The figures are illustrative examples of what a growth team might measure with Agent Serve, not customer results.
Last updated
Enterprise access
Agent Serve is available to teams by invitation. Tell us what you want to build and we will set up your workspace with you.