Actioning Facebook Ads

Actioning turns insights into changes you can push straight to Facebook (Meta) Ads — no copy‑pasting between Aliveo and Ads Manager. When you ask Aliveo to optimize budgets, bids, or statuses across your campaigns, ad sets, and ads, it produces an action table: a clear, side‑by‑side preview of current vs proposed values. You review the rows, adjust anything you like, select what you want, and click Apply — Aliveo then sends those changes to Meta on your behalf and reports back which ones succeeded.

This guide covers:

How action tables work — reviewing and applying changes safely.
What you can action — daily budgets, ad‑set bids, and statuses across campaigns, ad sets, and ads.
Creating actions on the fly with /actioning in Analytics Chat.
Running repeatable Facebook management agents for budget reallocation and pacing.
Example prompts, from simple one‑liners to multi‑step optimization routines.

Access note: Actioning is only available for accounts you have management access to. If you only have read access, Aliveo will still analyze and recommend, but the Apply controls won't appear.

1. The Action Table

Whenever a chat answer or an agent step proposes changes to your account, Aliveo renders them as an action table. This is the single place where you review and apply changes — it looks and behaves the same whether the table came from a chat, a /actioning request, or a scheduled agent.

Every action table carries the same instruction at the top:

Select rows, adjust the recommended values if needed, and click Apply.

Each row pairs a proposed value with its current value, and you can adjust any proposed value inline before applying. Select the rows you want, and the Apply button reflects how many you've chosen.

Anatomy of the table

Element	What it does
Row checkboxes	Select exactly which recommendations you want to apply. A header checkbox selects/deselects all eligible rows.
Proposed value (editable)	The recommended new value. You can edit it inline before applying — change a bid, round a budget, switch a status, etc.
Current value	Shown beneath the proposed value (e.g. Current: 2,000) so you always see what you're changing from.
Status badge	Pending until applied, then Applied once the change reaches Meta successfully.
Search / filter / sort	Narrow large tables with the Filter rows… box, per‑column filters, and sortable headers — handy when a table has hundreds of ad sets.

Proposed cells adapt to the field type: amounts like bids and budgets use a number input (with valid bounds), while statuses use a dropdown of allowed values.

Reviewing → applying

Review each row — compare the proposed value against the current one.
Adjust any proposed value inline if you'd like something different from what was recommended.
Select the rows you want with their checkboxes (or use the header checkbox to take everything).
Click Apply (N) — the number reflects how many rows you've selected. The button stays disabled until at least one row is selected.
Aliveo validates the changes, pushes them to Meta, and shows the result:
- Actions applied — "Selected recommendations were sent for execution."
- Some actions could not be applied — a few rows failed validation; review them and retry.
- Failed to apply actions — try again, or Refresh the table to reload the latest recommendations.

Applied rows flip to an Applied badge and their checkboxes lock, so you can't double‑apply the same change. You can keep working through the rest of the table, or hit Refresh to pull fresh recommendations against the now‑updated account.

Tip — expand for big tables. Use the maximize button to open the table full‑screen for easier review, and the refresh button to reload recommendations after the account changes.

Built‑in guardrails

Before anything reaches Meta, Aliveo runs each row through safety checks so you don't get cryptic API errors:

No‑op protection — rows where the proposed value equals the current value are skipped.
Drift protection — if the account changed since the recommendation was generated (the "current" value no longer matches Meta), the row is held back rather than overwriting newer data.
Meta rule enforcement — Aliveo pre‑checks Meta's constraints (see Things to know) and rejects rows that would violate them — e.g. a non‑positive budget or bid, or an unsupported status — with a plain‑language reason on the row.
Change log — every applied change is recorded (what changed, from → to, who, and when) for an auditable history.

2. What You Can Action

Aliveo can action the most common day‑to‑day optimizations across the Facebook Ads hierarchy — Campaigns → Ad Sets → Ads. Out of the box, actioning updates existing entities (it doesn't create or clone them — see Advanced).

Campaigns

Action	Values
Daily budget	Set a new daily spend cap (applies when the campaign holds the budget — i.e. Advantage Campaign Budget, formerly CBO)
Status	`ACTIVE` / `PAUSED`

Ad Sets

Action	Values
Daily budget	Set a new daily spend cap (applies when the budget lives at the ad‑set level)
Bid amount	Set the ad set's bid (the bid/cost control used by bid‑capped, cost‑cap, or ROAS‑goal strategies) — must be a positive amount
Status	`ACTIVE` / `PAUSED`

Ads

Action	Values
Status	`ACTIVE` / `PAUSED`

Where the budget lives. Meta keeps the budget at one level per campaign — either the campaign (Advantage Campaign Budget) or the ad set, never both. Aliveo proposes the budget change at whichever level your campaign actually uses; it doesn't move the budget between levels.

Where the bid lives. Meta exposes the bid at the ad‑set level only (there's no separate campaign‑ or ad‑level bid). A bid only applies when the ad set runs a bid‑controlled strategy — on lowest‑cost (no cap), there's no bid to set.

Note: Out of the box, actioning updates budgets, bids, and statuses on existing entities. Looking to also change targeting, clone winning ad sets/campaigns, switch bid strategies, manage lifetime budgets, or create entities from Aliveo? Talk to us — we can set these up as custom workflows for your account (see Advanced).

3. `/actioning` in Analytics Chat

The fastest way to make a one‑off change is right inside a chat. Type / in the chat input to open the mode menu, then pick actioning.

Selecting actioning tells Aliveo you want to change the account, not just analyze it. It then makes the actioning tools available so your request comes back as a ready‑to‑apply action table instead of a plain answer.

You don't always have to type /actioning — Aliveo also recognizes actioning intent from phrasing like "pause…", "reduce the daily budget…", "increase the bid…", or "turn off…". Using the slash command simply makes your intent explicit.

Things you can say in actioning mode:

"Pause all ad sets with zero purchases and more than $100 spend in the last 30 days."
"Reduce the daily budget by 20% for my two highest‑CPA campaigns."
"Increase the ad‑set bid by 15% on the 5 ad sets with the best ROAS."
"Pause the 3 ads with the highest frequency in my prospecting campaign."

Aliveo finds the relevant entities, builds the action table, and hands it to you to review and apply.

The other chat modes, for reference: info (answer from context without running analysis), brainstorming (strategy and market exploration), web_search (pull in live external info), and create_agent (draft an automation agent). Only actioning produces apply‑ready action tables.

4. Facebook Management Agents

For optimizations you want to run repeatedly — or schedule on a cadence — use a Facebook management agent instead of a one‑off chat. Two come ready‑made for the most common Meta optimization loops, and a third is a blank‑slate Custom Actioning Workflow you assemble yourself — each produces an action table you review and Apply just like in chat.

You'll find them in the agent creation flow under Step 2: Template Selection (in the Optimization Agents group), named Ads Management : Budget Reallocation, Ads Management : Pacing, and Custom Actioning Workflow.

Ads Management : Budget Reallocation

Reallocate ad‑set budgets based on CPA performance.

Looks at ad sets optimizing for conversions, benchmarks each ad set's CPA against a target you set, and proposes daily‑budget moves — raising spend on efficient ad sets and trimming the inefficient ones.

Runs on: Ad performance data, at the ad‑set level.
Default schedule: weekly.

Ads Management : Pacing

Bid change recommendation for effective pacing, with one‑click apply.

A two‑step routine that first assesses month‑to‑date pacing — total spend and CPA against your budget and CPA targets to decide whether the account is over‑ or under‑pacing — then proposes ad‑set bid adjustments to bring pacing back on track (for example, nudging bids up on under‑pacing ad sets that still have CPA headroom, capped to a sensible step).

Runs on: Ad performance data, at the ad‑set level.
Default schedule: one‑off (so you stay in control of when bids change).

Custom Actioning Workflow

Blank‑slate actioning workflow for Facebook Ads at the campaign / ad‑set / ad level (budget, bid, status).

When neither ready‑made agent fits, start from this blank‑slate template — it drops you into the playground with the actioning tools already switched on, so you assemble the analysis and the change yourself.

Runs on: Ad performance data.
Default schedule: one‑off (so you stay in control of when changes are proposed).

Building your own

All three follow the standard agent‑building flow, so you can clone a ready‑made agent and adjust its thresholds, time windows, and rules — or start from the Custom Actioning Workflow and assemble a fresh routine from scratch:

Shared Context — define anything common across steps (definitions, thresholds, the time window, the CPA/spend targets).
Analyze steps — do the data work: pull performance, compute metrics like CPA/ROAS/CTR, rank entities, and produce the proposed changes as an action table.
Summary step — (optional, always last) logically summarize the combined output of the analyze steps.

Because these are actioning agents, the analyze step that proposes changes renders an action table in the output — you still review and Apply exactly as you would in chat.

For the general agent‑building mechanics — accounts, data context, grouping, steps, preview, and scheduling — see Creating an Agent.

5. Examples — From Simple to Complex

Every example below is a single natural‑language request. Aliveo handles the analysis (finding the entities, computing the metric, ranking) and produces the action table — you review and apply.

Simple, single‑step actions

These map directly to one type of change:

Budget — "Set the daily budget to $200 for my top 2 campaigns by purchases."
Status (ad set) — "Identify the 2 ad sets with the highest spend and pause them."
Status (ad) — "Pause the 3 ads with the most spend and no purchases in the last 14 days."
Bid — "Find the 2 ad sets with CPA above $40 and reduce their bid by 25%."

Metric‑driven optimizations

These combine an analysis with a change — Aliveo computes the metric, ranks, filters, then proposes the action:

CPA budget trim — "Find the 2 campaigns with the highest CPA and reduce their daily budget by 20%."
Wasted‑spend cleanup — "Identify 3 ad sets with zero purchases and spend above $100 in the last 30 days, and pause them."
Bid efficiency — "For the 2 ad sets with the worst ROAS, reduce the bid by 15%."

Pause underperformers, scale winners

A pair of complementary moves you can run together or on a schedule:

Cut the losers — "Pause ad sets where CPA is more than 1.5× the account average over the last 30 days." / "Pause ads with a CTR below 0.5% and more than $50 spend."
Scale the winners — "Increase the daily budget by 20% on the top 3 ad sets by ROAS." / "Raise the bid by 10% on ad sets beating their CPA target."

Watch the learning phase. Meta can re‑enter an ad set into its learning phase after a large budget or bid change (roughly >20%). The management agents keep step changes moderate for this reason — if you edit a proposed value by hand, keep your steps small to avoid resetting learning.

Creative fatigue management

Turn frequency and engagement trends into ad‑level cleanup:

High frequency — "Pause ads with a frequency above 3 and declining CTR week‑over‑week."
Worn‑out creative — "Identify the 3 ads with the steepest CTR drop in the last 7 days and pause them."

Pacing & budget reallocation

Reallocate — "Across my conversion ad sets, raise the daily budget 15% on those beating their CPA target and cut it 20% on those above target." (or run the Budget Reallocation agent)
Pace to target — "We're under‑pacing this month — raise bids on under‑pacing ad sets that still have CPA headroom, capped at +25%." (or run the Pacing agent)

Multi‑step routines you can assemble

In a management agent, you can chain several of the above into one routine. A few patterns that work well:

Weekly account hygiene — (1) Analyze: find zero‑purchase, high‑spend ad sets → pause them. (2) Analyze: find high‑frequency, fatigued ads → pause them. (3) Analyze: rank remaining ad sets by ROAS → raise budget on the best. (4) Summarize what changed.
Budget & pacing loop — (1) Analyze: assess month‑to‑date pacing vs target. (2) Analyze: reallocate ad‑set budgets by CPA. (3) Analyze: adjust ad‑set bids to correct pacing. (4) Summarize.

Each analyze step produces its own action table, so you can review and apply each set of changes independently.

Advanced, tailored workflows

Beyond the self‑serve actions above, we build custom actioning workflows tuned to your account's strategy and scale. Several capabilities live here today — if any sound like what you need, talk to us and we'll set them up for you:

Targeting changes — adjust audiences, geos, age/gender, placements, or device targeting on existing ad sets as part of a workflow.
Cloning & duplication — duplicate winning ad sets or campaigns into new audiences, geos, or budgets to scale what's working.
Bid‑strategy switching — move ad sets between lowest‑cost, bid cap, cost cap, and ROAS‑goal strategies.
Lifetime budgets & spend caps — manage lifetime budgets, campaign spend caps, and CBO/Advantage Campaign Budget toggles (beyond the daily‑budget self‑serve actions).
Dayparting & scheduling — schedule ad sets to run at specific times/days.
Audience operations — expand or exclude audiences, optimize placements, and consolidate fragmented ad sets to exit the learning phase faster.
Sophisticated bid & budget management — rule‑based and goal‑seeking logic that reacts to efficiency, ROAS, and conversion trends rather than simple top‑N rules.
Creating entities — spin up new campaigns, ad sets, or ads from Aliveo.

These run as the same review‑and‑apply action tables, so even sophisticated logic stays fully under your control — nothing changes until you click Apply.

6. Things to Know — Meta's Rules

Aliveo enforces Meta's own constraints up front, so invalid rows are caught before they're sent. If you see a row rejected with a reason, it's usually one of these:

Budgets

Budget lives at one level — either the campaign (Advantage Campaign Budget) or the ad set, never both. Aliveo proposes the change at whichever level holds the budget and won't move it between levels.
Daily budgets only via self‑serve actioning — lifetime budgets and spend caps are handled as custom workflows.
Budgets must be a positive amount and meet Meta's minimum daily budget for the ad set's optimization/billing setup.

Bids

Bids are set at the ad‑set level — there's no campaign‑ or ad‑level bid to change.
A bid only applies on bid‑controlled strategies (bid cap / cost cap / ROAS goal). On lowest‑cost (no cap), there's no bid to set.
Bid must be a positive amount.

Status

Only ACTIVE / PAUSED can be set. Archived or deleted entities can't be re‑activated through actioning. Aliveo normalizes phrasing like "enable", "disable", or "turn off" to the right value automatically.

Learning phase (good to know)

Significant budget or bid changes can push an ad set back into Meta's learning phase, which temporarily affects delivery. This isn't an error — but it's why the management agents favor moderate step changes, and why you should keep manual edits small.

These guardrails mean you can safely propose broad changes; Aliveo simply skips the rows Meta wouldn't accept and tells you why.

That's actioning — analyze, review, and apply Facebook Ads changes without leaving Aliveo. Start with a quick /actioning request in chat, and graduate to a Budget Reallocation or Pacing agent once you find an optimization worth repeating.