Skip to main content
Tracking Health shows whether the events your pixels record are reaching the platforms you connected, how clean your traffic looks, and how well Meta can match your events. SignalBridge also emails you when something needs attention. The dashboard looks at the last 24 hours and refreshes every few minutes while the page is open.

What It Monitors

Platform Delivery

Delivery to Facebook CAPI, Google Ads, TikTok, Taboola, Microsoft Ads, Klaviyo and Google Analytics 4, each measured separately.

Traffic Quality

An independent rating of how much of your traffic looks suspicious, with the signals behind it.

Event Match Quality

Your Meta Event Match Quality (EMQ) score per event type, with the opportunities that would raise it.

Proactive Alerts

Email alerts when delivery breaks, a platform goes quiet, tracking stops, or match quality declines.

Reading the Dashboard

The summary banner

The banner shows the overall state (Delivery operational, degraded, or issues detected), the reasons when it is not healthy, and two figures:
  • events — how many events your pixels recorded in the last 24 hours.
  • ingested — the share of those events SignalBridge accepted and processed. This is about what reaches SignalBridge. It is not a delivery rate and says nothing about what each platform received.
Choose a pixel from the selector to look at one pixel, or leave it on All Pixels to combine them. A pixel with no events and nothing to report shows No data rather than a status.

Platform Delivery

Each connected platform gets a row with its delivery rate and the number of events delivered out of the number attempted. A few things you may notice on a row:
  • Action needed — you have something to do, usually reconnecting the integration because the saved authorization expired or was revoked. Open the row for the details and a link to the integration settings.
  • N rejected after retries — events the platform refused and SignalBridge stopped retrying. The status is unchanged while the delivery rate is healthy, but the count is shown so nothing is hidden. Open the row for the reason.
  • quiet for Nh — the pixel is still recording activity, but no events have reached this platform for several hours. It is informational: a paused campaign or a low-volume audience can look the same.
  • Delivery data temporarily unavailable — SignalBridge could not read delivery data for that platform just now. The other platforms are shown normally and the row refreshes on its own.
Select any row to open its detail panel: the current status, what SignalBridge is retrying on your behalf, and the technical evidence behind any issue.

What each rate measures

Rates are not identical across platforms, because platforms confirm delivery differently.
  • Facebook CAPI, Google Ads, TikTok and Taboola — events the platform accepted, out of events attempted.
  • Microsoft Ads and Klaviyo — events the provider’s API accepted. Acceptance is not attribution: an accepted Microsoft upload can still be reported as unattributed in your Microsoft account.
  • Google Analytics 4 — HTTP receipts. GA4 confirms that its endpoint received a request; it does not report per-event processing or attribution, so this percentage measures transport delivery only. Events not sent because of marketing consent, or too old to send, are excluded rather than counted as failures.

Traffic Quality

Traffic Quality is reported independently of delivery. A pixel can deliver every event perfectly while some of its traffic is suspicious, and the reverse. SignalBridge classifies each eligible event by the visitor’s network and behavior. The rating is the share of classified events that look suspicious. Verified search and social crawlers, such as Googlebot, are blocked before your analytics and ad platforms see them, so they are not counted in the rating. The panel shows how many were blocked: A rating is only shown when there is enough evidence: at least 25 classified events, and at least 80% of eligible events classified. Otherwise you see Insufficient data or Limited coverage, together with how many events could not be classified. The signals shown beneath the rating explain why events were flagged: VPN, proxy, Tor, datacenter, hosting, relay, malicious, rate anomaly, bot and blocklist. One event can carry several signals, so they can add up to more than the suspicious total. Bot covers other blocked bots, such as automated scripts, and never includes verified crawlers.
Some real visitors are flagged: people on a VPN, iCloud Private Relay, or a shared office or mobile network. Review the signals before changing any settings. The Review in Ad Shield link on a Fair or Poor rating opens Ad Shield, where you can see the most suspicious IP addresses and adjust the detection thresholds for each pixel. Ad Shield reports on traffic and estimates wasted spend; it does not block visitors. By default it only observes. For each pixel you can opt in to keeping high-confidence fraudulent traffic out of ad platforms, after a preview of exactly what would be withheld; those events are still recorded and counted in SignalBridge.
If the classification pipeline is behind, the panel says Classification delayed and recent traffic may be missing from the figures.

Email Alerts

The alert check runs every 30 minutes. Most alerts wait for two consecutive checks before sending, so expect an email roughly 30 to 60 minutes after a problem starts. A few serious conditions, such as rejected events or a very high failure rate, send after the first check. SignalBridge emails you when:
  • Events stop — no events were received in 24 hours on a pixel that normally records activity.
  • Traffic falls far below normal for the same weekday and hour.
  • Events fail — the ingestion failure rate is above 5% (critical above 15%).
  • A platform rejects events after retries, or a backlog of retries has been waiting for more than an hour.
  • Delivery is degraded — under 80% of attempts to a platform succeeded.
  • A platform goes quiet — it delivered at least 100 events earlier in the last 24 hours, then received none for at least 12 hours while your pixel kept recording activity.
  • A connection needs attention — an integration must be re-authorized.
  • A conversion mapping is missing — paid-click events are arriving but no offline conversion is mapped for Google Ads or Microsoft Ads.
  • Shopify browser tracking may be disabled — your store reports orders that the browser pixel never saw.
  • Meta EMQ declines significantly.
When a problem clears, SignalBridge sends a Tracking restored email. An incident about rejected events or a retry backlog also closes on its own, without an email, once the last 24 hours contain no rejected events, no retries and no delivery attempts for that platform. Alerts for a platform you connected within the last seven days are held back for delivery-rate and quiet-platform checks, so a new integration has time to build a representative sample. Configure alert preferences in Settings → Notifications.

Accessing Tracking Health

Open Dashboard → Overview to see Tracking Health at the top of your dashboard.