Home
Home is your decision queue. Not a wall of charts to interpret — a ranked list of things worth deciding today, each with the money at stake and the reasoning attached, and a button to act.
In one line
Work top to bottom: money at risk first, then growth headroom, then check whether last week’s decisions paid off. Most people clear it in a few minutes each morning.
The headline#
At the top: how much needs a decision, split into what’s at risk and what’s available as headroom, and what share of your trailing spend that represents. It decomposes exactly into the sections below it, so you can always trace the number to the rows that made it.
Before there’s a queue to show, the headline says so honestly instead of claiming you’re all set — “pulling in your account data” while the first sync runs, “the first check is queued” once your data has landed. Adlyse checks your accounts after each sync rather than on a fixed clock, so the first proposals arrive when there’s something real to say.
The sections#
Each one holds a different kind of decision, ordered by what usually matters most.
Money at risk#
Spend that isn’t earning its keep — a campaign burning budget with no conversions, a cost per result far above the account’s norm, a broken conversion tag. These are the rows that pay for the subscription.
Growth headroom#
The opposite: places you’re capped by budget rather than by demand, where more spend should convert at roughly the same efficiency. Approving one of these raises a daily budget.
Results came back#
Decisions you already approved, now old enough to judge. Each shows what was predicted against what actually happened — win, below, inconclusive, or still scoring. This is the section that makes the queue trustworthy over time: you can see whether Adlyse’s estimates hold up in your account.
Experiments#
Tests currently running, and the ones that have concluded and need a call.
Strategy#
Bigger proposals about direction rather than individual campaigns.
Needs you#
Things Adlyse found but can’t act on — a missing permission, a disconnected account, a conversion action that isn’t set up. Each names what’s blocked and how much it’s holding back. Worth clearing early, because everything downstream is limited by these.
Other proposals#
Everything else worth a look, below the ranked sections.
Working a decision#
Click any row to open it. You get the reasoning, the evidence behind it, and its estimated monthly impact with a confidence level.
From there you can:
- Approve it — you’ll see a before-and-after preview of the exact change before anything is applied.
- Dismiss it, with a reason.
- Snooze it, if it’s real but not now. Snoozed rows collect in their own section rather than vanishing.
- Ask about it — open Adlyse AI on that specific row: “why are you recommending this?”, “which should I do first?”, “what happens if I skip it?”
Nothing is applied without you
Every proposal is pending until a human approves it. This holds for proposals written by scheduled agents too — an unattended run can’t change your ad platforms directly, so it leaves you a decision instead. That’s why the queue is the destination for automation, not a bypass around it.
What I did, and whether it worked#
Below the queue, the outcome ledger keeps the running history: what you approved, what it predicted, and what actually happened. Dismissed rows are kept too, with the reason. It’s the fastest answer to “did we get value out of this last quarter” — and the honest record you’d otherwise have to rebuild from memory.
Filtering and freshness#
Switch between the whole organization and a single ad account with the account filter. The header tells you how current the underlying data is, so you know whether you’re looking at this morning’s picture or yesterday’s.
When the queue is empty#
An empty queue is the goal, not a bug. When there’s nothing to decide, Home shows what’s worth doing on an account that’s healthy today — plus tips relevant to your connected platforms.
Common flows#
- The morning pass. Clear Money at risk, glance at Growth headroom, done.
- Prove the value. Open Results came back before a client call and read the wins out.
- Unblock yourself. Work Needs you once a week; those rows cap everything else.
- Argue with it. Ask Adlyse AI to walk you through the evidence before approving anything large.
Drill from a proposal into the campaigns, ad sets, and ads behind it.
Ask Adlyse AIInterrogate any row before you approve it.
Feed the queue automaticallyScheduled agents propose here while you’re not looking.
ReportsTurn a period of decisions into something a client can read.
Pull the same queue as JSON with list_suggestions and get_suggestion, and the account-level snapshot with get_dashboard_snapshot.