Monitor campaigns, prices, visibility and product data. Turn defined conditions into tasks, alerts or configured actions.
No account needed: see the detected deviation, evidence and next step in a sample task.
Monitoring becomes useful when a event has a defined meaning and someone knows what to do next. Choose the scope, establish the condition and connect the result with a task, notification or supported action.
Choose what you want to monitor: a product, campaign, destination URL or tracked phrase. Set the relevant period and threshold. Compare budget pace with a plan, and allow purchase data time to mature before judging a product’s latest results.
Open the task to inspect the object, observed change and available history. Check how fresh and complete the data is. A failed URL reading, a price mismatch and falling impressions need different follow-up checks, even if all require attention.
Assign responsibility for the next step. Depending on configuration, a process can create a task, send email or SMS, make a Twilio call, or run a supported action. A subsequent reading helps establish whether the problem remains.
Choose an account, products and process conditions. Set a tolerance, reference period and schedule appropriate for the situation.
A task should identify the source, period and object behind the observation. Review data freshness and completeness before making a decision.
Send an alert to the right person, create a task or run a supported automation. Later readings help establish whether the situation was resolved.
Use the price view to compare your offer with competitors and inspect readings for the same product over time. Assess changes alongside availability and business assumptions. Keep this separate from a mismatch between feed and store prices, which calls for a data check.
Inspect the position history of a tracked phrase or the impressions of a selected product. Each view has its own scope and period. Open the related data in Sembot from a task, then return to the investigation; ranking and delivery measure different things.
A forecast moving away from plan calls for a pacing review. For a broken destination, establish whether the error repeats and which ads still use it. Traffic cost over a reporting period is not automatically a confirmed loss during the outage.
No. Start with a task and notification while leaving the decision to a person. Automatic execution applies to supported actions and requires suitable configuration and permissions; it is not the default response to every observation.
Choose a meaningful scope, a sensible threshold and a person responsible for responding. Match the channel to the urgency. A task should provide enough context for the next step, rather than simply announce that a metric has changed.
It is not confirmation that everything is working. Check the source and data freshness before closing the task. A “missing data” result needs further information rather than being treated as a successful check.
Choose a process that matches a specific problem you want to monitor.
Set the scope, period, threshold and notification channel for the responsible person.
Walk through an example task and identify the reading that would confirm resolution.
Choose a process, connect its source and review the first result. Find package and add-on details in pricing.
