Ad accounts fail quietly. A campaign starts overspending at two in the morning, a best seller stops converting because the listing changed, a profitable campaign runs out of budget every afternoon for a month. By the time anyone looks at the weekly report, the money is gone.
Alerts fix that, but only if there are few of them and each one means something. This playbook covers the five alerts worth setting on almost any account, what each one catches, how to set thresholds that do not cry wolf, and what to do when each fires.
What makes an alert worth having
A good alert meets three tests. It catches a problem that costs real money if left alone. It fires rarely enough that you take it seriously. And it comes with a clear next step. An alert that fires daily and leads to no action trains you to ignore all of them.
The most reliable way to meet those tests is to compare each campaign to its own history, not to fixed numbers. A spend level that is normal for one campaign is a spike for another.
1. Spend spike
What it catches: a campaign or account spending far faster than usual, often from a bid change, a broad match term suddenly matching a high-volume search, or an automated change gone wrong.
How to set it: compare spend so far today against the trailing average for the same time of day. Alert when today's pace is well above normal, for example half again above the average by midday. Ignore campaigns with very small budgets.
What to do: check the search terms and recent changes on the campaign. If the cause is not obvious, reduce the budget until it is. The spend spike rule covers thresholds in detail.
2. Profitable campaign out of budget early
What it catches: a campaign that converts well but runs out of budget before its best hours. This one costs sales rather than wasting spend, so it rarely shows up as a problem in reports.
How to set it: alert when a campaign below your target ACoS hits its daily budget before a set hour, such as mid-afternoon, several days in a row.
What to do: raise the budget, or lower bids or pause weak hours so the same budget lasts into the evening. Budget exhausted early walks through both fixes.
3. ACoS creep
What it catches: efficiency drifting slowly in the wrong direction. Day to day, the change is invisible. Over a month, it can turn a profitable campaign into a loss.
How to set it: compare the trailing seven-day ACoS against the trailing thirty-day ACoS for each campaign with meaningful spend. Alert when the short window is clearly worse than the long one, and above your target. Using trailing windows smooths out daily noise and late attribution.
What to do: look for the cause before cutting bids. Common causes are rising CPCs, new competitors, a price change or a listing change. ACoS creep detection covers the diagnosis.
4. Delivery stopped
What it catches: a campaign that normally spends every day and suddenly spends nothing. Causes include a payment problem, a suppressed or out-of-stock listing, a lost buy box, or a campaign paused by mistake.
How to set it: alert when a campaign with steady daily spend records little or none for a full day, or by a set hour.
What to do: check the listing first, then billing, then campaign status. This alert often catches problems outside advertising, which is part of its value. A suppressed listing hurts organic sales too.
5. Conversion drop with steady spend
What it catches: clicks still arriving at the usual cost, but orders falling away. That points at the listing or the offer, not the ads: a price increase, a lost buy box, a new negative review, a broken image or a competitor undercutting you.
How to set it: compare conversion rate over the last few days against the trailing average for campaigns with enough clicks to be meaningful. Alert on a sharp fall while spend is roughly unchanged.
What to do: look at the product page as a shopper would. Fixing the listing is usually the answer. Cutting bids only hides the symptom.
Alerts that usually are not worth it
Daily ACoS above target. Daily ACoS swings with attribution delay and small numbers. It fires constantly and teaches you to ignore alerts.
Every keyword with clicks and no sales. That is a weekly review task, not an alert.
Small changes in impressions or CTR. Useful in analysis, rarely urgent.
Keep these in the weekly routine instead. The monitoring checklist sorts daily, weekly and monthly checks.
Notify or act?
Some alerts can safely take action on their own. A spend spike on a campaign with a hard cap can reduce budget or pause automatically, because the downside of waiting is larger than the downside of a brief pause. The others usually need a person to find the cause first. For those, a clear notification with the campaign, the numbers and a link is the right output. How spend alerts work covers the setup.
Frequently asked questions
What alerts should I set for Amazon ads?
Start with five: a spend spike against your normal daily pace, a profitable campaign running out of budget early, ACoS drifting up over several weeks, a campaign that stops spending unexpectedly, and conversions falling while spend holds steady. Together they cover the most expensive problems.
How do I avoid alert fatigue?
Set thresholds relative to each campaign's own history rather than fixed numbers, alert only on campaigns that spend enough to matter, and give each alert a clear action. If an alert fires often and you never act on it, raise its threshold or remove it.
Should alerts take action automatically?
Some can. A spend spike on a campaign with a clear cap can safely pause or reduce budget automatically. Alerts about conversion drops or delivery stopping usually need a person to diagnose the cause first, so a notification is better than an automatic change.
Off Hours includes spend alerts free on every plan and adds performance rules for ACoS and conversion checks, all running on a 15-minute cadence. Start a free 14-day trial.