Alli Alerts is marketing anomaly detection and campaign monitoring for every channel. It combines data science models with custom alert rules to catch performance outliers, budget pacing issues, broken URLs, feed failures, and auction shifts, then notifies your team in Slack so they can act early.
higher cost effectiveness than benchmark
video completion rate
ad versions sliced from 7 videos
Alli Creative Insights scored every Sterling ad version, and Alli Alerts flagged pacing issues so spend could move to what worked. Read the Sterling customer story
Marketing anomaly detection uses statistical models to spot when a campaign metric moves outside its normal pattern, like a sudden drop in conversions, a spike in cost, or a surge in clicks. The system flags the change as soon as the data shows it, so nobody has to find it in a report.
The day starts with logging in to check that nothing broke, budgets are pacing, and feeds are healthy. That hunt eats the hours meant for strategy.
A campaign that stops spending on Friday can stay dark all weekend, and a broken landing page keeps charging for clicks until someone notices.
Simple percentage alerts trip on normal weekday swings. After a few false alarms, teams start ignoring them, including the one that mattered.
Standard Alerts start from a PMG global baseline. Custom Alerts add the rules your business cares about. Both watch every connected channel from one alerting layer.
Alli Anomaly Detection uses models built by PMG’s data science team. Each metric is judged against its own history, including weekday swings and seasonality, so normal volatility stays quiet.
For your teamYou catch the drop, or the breakout, while there’s still time to act.
Build triggers for a client, channel, or market, from Merchant Center feed status to market-level pacing to keyword-level URL issues. SQL views and Marketplace apps handle more advanced logic.
For your teamThe checks your team still does by hand become rules that run on their own.
Explore Alli MarketplaceBudget caps, no-spend checks, hourly pacing, URL scraping, feed monitoring, ad strength, and auction signals all sit in one list, running on the same governed data as every report.
For your teamQuality holds across thousands of campaigns without anyone checking each one.
No-spend issues, overspend, pacing problems, and abnormal spend, including fraudulent activity, are flagged the moment they appear, overnight and on weekends too, so teams can act before costs build up.
Feed failures, out-of-stock products, and broken landing pages get caught before they disrupt sales, especially in peak periods.
Brand impression share drops and new advertisers in your auction show up while there’s still time to respond.
Repetitive monitoring turns into time spent on strategy and optimization, without adding headcount to watch dashboards.
Who uses it
Skip the morning platform checks.
The system watches every platform, so buyers respond to the issues that need them.
Protect peak shopping windows.
Feed failures and out-of-stock products still serving get flagged before the busiest days of the year.
See auction pressure as it builds.
Brand impression share drops and new competitors entering the auction arrive while bids can still change.
Know spend is protected.
Quality standards like ad strength and sitelinks hold across every account, and issues surface fast.
An alert is the start of a decision. Alli Alerts delivers it where the team already works, organized so the right owner picks it up.
Set when each alert runs and who hears about it. Alerts arrive in Slack, where media, analytics, operations, and client teams can share and triage them together.
For your teamNobody has to log in to another dashboard to find out something broke.
Labels group alerts by team and purpose. For a no-spend alert, check feed health in Datasources first. If conversions drop while traffic holds, check landing pages and tracking before bids.
For your teamTeams fix the cause on the first try, and a data issue doesn’t get treated as a media one.
See connection health in DatasourcesAlli Alerts is a two-layer monitoring system: Standard Alerts with PMG anomaly detection, and Custom Alerts for your own rules.
Alerts runs on the data Datasources brings into Alli.
Anomaly detection, guardrails, and hygiene checks start from a PMG global baseline.
Build triggers for your own thresholds, with SQL views and Marketplace apps where needed.
Alerts arrive in Slack, so the right people see them and triage together.
Investigate in Data Explorer, fix it in the platform, or start a follow-up in Generative Workflows.
Six kinds of alerts cover performance, spend, site and feed health, account quality, and the auction. Custom Alerts add your own rules on top.
Have a check your team still runs by hand? It can usually become a custom alert. Ask about alert setup
Manual checks depend on someone looking at the right thing at the right time. Native alerts only see their own platform, each with its own logic.
| Manual monitoring | Native platform alerts | Alli Alerts | |
|---|---|---|---|
| Coverage | Whatever someone has time to check | One platform at a time | Every connected channel, in one alerting layer |
| Detection logic | Human judgment | Each platform’s own rules | Data science anomaly detection plus custom rules |
| Noise | Depends on the person | Simple thresholds that trip on normal swings | Models that account for historical patterns |
| Hygiene checks | Spot checks | Limited | Automated URL scraping and feed monitoring |
| Competitive signals | Found in reports later | Limited | Impression share drops and new auction entrants |
| Delivery | Whoever noticed tells someone | Each platform’s inbox | Slack and the workflows your team uses |
PMG manages Alli Alerts as a global operating standard, giving every client a consistent baseline for monitoring campaign health, investment risk, and data integrity.
Security: the Alli platform holds a current SOC 2 Type II report and is aligned with the NIST Cybersecurity Framework (CSF). Request the report See security details
Standard Alerts apply the same checks across every market and account, and teams add custom rules on top.
The same detection logic runs across every connected platform, so problems that span channels get seen.
Alerts read the same governed data as every dashboard, defined once in the Semantic Layer.
The ideas behind Alli Alerts
What Alli Alerts monitors, how it avoids noise, and how teams use it.
Alli Alerts is marketing anomaly detection and campaign monitoring inside Alli. It combines Standard Alerts, including PMG’s data science anomaly detection, with Custom Alerts for client and channel rules. It monitors performance, budget pacing, URLs, product feeds, ad quality, and auction shifts, and delivers alerts into Slack.
A threshold alert fires whenever a metric moves a fixed amount, like spend dropping 20%, so it trips on normal volatility. Alli Anomaly Detection uses models that learn each metric’s historical patterns, so it flags changes that are unusual for that metric and ignores routine swings. That means fewer false alarms.
Thresholds and conditions are set together with your team, around signals that need a decision. You define the acceptable range and the conditions that trigger a notification, and model-based anomaly detection filters out normal volatility. The team hears about what matters, and small day-to-day movements stay quiet.
Alerts are delivered into the communication tools your team already uses, including Slack. That makes them easy to share, triage, and act on across media, analytics, operations, and client teams, without anyone needing to log in to another dashboard to see them.
Alli Alerts watches your signals every minute of every day, checking data as it arrives, and investment guardrails support hourly pacing triggers. How fast an alert fires depends on how often each platform reports its data. For most issues, teams hear about them well before the next weekly report.
Native alerts only see their own platform, each with its own logic. Alli Alerts creates one alerting layer across every connected channel, with consistent detection logic, data science anomaly detection, hygiene checks, and competitive signals, so teams get a cross-platform view in one place.
Yes. Custom Alerts let practitioners build triggers for specific client, channel, or market needs, such as budget capping, Merchant Center feed status, no-spend alerts, or market-level pacing. Advanced teams can use SQL views and Marketplace apps to build more complex monitoring logic.
Start with the alerts that protect money: no-spend, budget caps, and broken landing pages on your highest-spend campaigns. Once the team trusts those, add anomaly detection and competitive alerts, then build custom alerts for the checks your team still does by hand.
Book an intro and we’ll show you the alerts your team would have received last month, and how much manual checking they’d have saved.