Skip to main content
Metrics type: Key MetricsCategory: Ad Platform

At a glance

The alert that fires when the AdRoll Pixel stops firing or conversion events collapse while campaigns keep spending. The tag fails quietly, purchase and add-to-cart events stop reaching NextRoll, attribution breaks, and the retargeting and dynamic-ad audiences that depend on fresh visitor signals begin to starve. It watches for clicks continuing on the normal pattern while confirmed conversions fall well below the level you would expect for that click volume, or for the live conversion-event rate dropping below a healthy threshold against the recent baseline. Because AdRoll is retargeting-first, a Pixel break does double damage: it breaks measured ROAS and CPA, and it stops feeding the visitor and cart-abandoner pools that most of your spend targets. The goal is to catch the break inside the same day, before a full cycle of spend reports as wasted and before the retargeting audiences shrink past the point of easy recovery. Caveat: a tracking break and a genuine demand drop look identical on a ROAS card alone, so this card exists to tell them apart.

Calculation

Calculated automatically from your AdRoll data. See the At a glance summary above for what the metric tracks and the worked example below for a typical reading.

Worked example

Northwind Outdoors, a UK outdoor-gear retailer on BigCommerce, runs AdRoll retargeting plus a small prospecting layer, with Dynamic Ads driven by their BigCommerce product feed. A tag-manager change at 09:00 on 14 Apr 26 reordered the container so the AdRoll Pixel loaded after a script that threw an error, and the Pixel stopped firing on most pages. All figures are illustrative.
  1. The alert fires at 09:40, about 40 minutes after the break. Clicks held on the normal weekday-morning pattern, but observed conversions fell well below half of expected and the event firing rate dropped from 97% to 38%. That divergence is the tracking-break signature.
  2. The event-rate collapse is the tell. A pure demand drop would lower conversions while the Pixel kept firing at its usual rate. Here the firing rate cratered, which means the tag itself stopped, not that shoppers stopped buying.
  3. Every hour of delay costs measured ROAS and audience freshness. AdRoll keeps spending while it can no longer record conversions, so the ROAS and CPA cards drift toward false alarms. Worse, the visitor and cart-abandoner segments stop receiving new members, so the retargeting pools begin to shrink and age.
  4. The fix is an engineering action, not a marketing one. Restore the tag-container order or re-add the Pixel snippet, then confirm in the AdRoll dashboard (Pixel section) that segment and conversion events are arriving again.
  5. Recovery is not instant. Once the Pixel is firing, attribution resumes immediately, but the retargeting pools take time to refill to their previous size, and any optimisation that throttled during the blind period needs a short window to recalibrate.
Quick reads:
  • Clicks steady + conversions near zero + event rate collapsed = tracking break. Engineering fix.
  • Clicks down + conversions down + event rate steady = genuine demand drop, not this card.
  • Event rate drifting down over days = creeping implementation drift (consent change, script reorder, catalogue tag removed). Audit before it becomes a hard break.

Sibling cards merchants should reference together

Reconciling against the AdRoll dashboard

Where to look in the AdRoll dashboard: The authoritative place to confirm a break is the AdRoll dashboard at app.adroll.com. Open the Pixel area (Settings, then the Pixel and conversion section) to see live event volume and whether your segments are still receiving members. The conversion-events view shows whether purchase and other conversion events are arriving, and the Pixel status indicator flags when AdRoll has not seen the tag fire recently. The campaign Reporting tab shows the downstream effect (conversions, ROAS and CPA moving) but not the cause; the Pixel view shows the cause. Other AdRoll views that look related but are not:
  • Attribution settings: configures the click and view windows (default 30-day click, 1-day view). Useful context but not a live health monitor.
  • Audience segment sizes: a slow-moving count. A shrinking segment is a lagging symptom of a Pixel break, not an early detector, because pools deflate gradually over days.
Why our number may legitimately differ from AdRoll: Cross-connector reconciliation:

Known limitations / merchant FAQs

How is this different from a normal conversion drop? A demand drop lowers conversions while the Pixel keeps firing at its usual rate and clicks fall alongside. A tracking break lowers measured conversions while clicks hold steady and the event-firing rate collapses. This card is tuned to the second pattern. If conversions and clicks fall together, look at Conversion Drop Alert instead. What usually causes a break? The common culprits are a theme or template deploy that removes or reorders the AdRoll Pixel snippet, a tag-manager container change that loads the Pixel after a failing script, a consent-banner change that blocks the Pixel until acceptance, a checkout-flow change that skips the conversion event, or a CMS migration that drops the snippet from key page templates. After any of these, expect this card to be the first to react. Why does it watch the event-firing rate as well as conversions? Because a partial break is the hardest to spot manually. If the Pixel fires on most pages but stops on the confirmation page, conversions can crater while overall traffic events look fine, or the reverse can happen. Watching the firing rate against its baseline catches the partial case that a raw conversion count alone would miss, and it reacts faster because segment events are higher volume than conversion events. Why does a Pixel break hurt more on AdRoll than on a search platform? Because AdRoll is retargeting-first. Most spend targets people the Pixel has already seen (site visitors and cart abandoners), and Dynamic Ads serve products the Pixel recorded a visitor viewing. A break therefore stops feeding the audiences you are actively spending against, so you lose both your scorecard and your targeting fuel. On a prospecting-only or search account, the audience-starvation half of the damage does not apply. Can a deploy trip this even though nothing is actually broken? Occasionally. A consent-mode change that delays the Pixel until after acceptance can briefly look like a break before events catch up, and a short CDN or gateway blip can drop a few minutes of events. The 30-minute sustained condition is there to filter transient blips, but a confirmed fire still deserves a check in the AdRoll Pixel view. How fast can I expect to be alerted? Within minutes of a sustained divergence, well inside the same reporting day. That is the whole point: catching it before a full cycle of spend reports as wasted, before the ROAS and CPA cards raise false alarms, and before the retargeting pools have meaningfully deflated. After I fix it, why do my retargeting audiences take time to recover? Attribution resumes the moment the Pixel fires again, but the visitor and cart-abandoner segments only refill as new shoppers browse, so a pool that drained during a long break needs days of normal traffic to return to size. Confirm the fix in the AdRoll Pixel view (events arriving and segments growing again) rather than waiting on the campaign ROAS number to fully normalise. Can I tune the thresholds? Yes. The expected-conversion ratio and the firing-rate floor are both configurable per profile in the Alert Rules tab. Tighten them on high-spend retargeting accounts where even a short break is costly; loosen them on low-volume accounts where small samples are noisy.

Tracked live in Vortex IQ Nerve Centre

AdRoll Pixel Tracking Broken is one of hundreds of KPI pulses Vortex IQ tracks across AdRoll and 70+ other ecommerce connectors. Nerve Centre runs the detection layer; Vortex Mind investigates the cause when something moves; Ask Viq lets you interrogate any number in plain English. Start for free or book a demo to see this metric running on your own data.