How we catch issues before they escalate
Ads, traffic, conversions, spend, SEO, and revenue are monitored continuously. Unusual changes are detected automatically and your team is alerted before problems become expensive.
Free preview. No credit card required.
From connected account to real-time protection
Connect your platforms
Click to connect your data in seconds. No code or technical setup required.
We learn what matters
Your data is analysed to identify the signals most important to your performance.
Baselines adapt automatically
Each metric builds its own behavioural baseline using historical patterns, trends, seasonality, and normal fluctuations. No rules or thresholds to configure manually.
Your team gets alerted instantly
The moment something unusual happens, alerts are delivered through Slack or email with context and likely causes.
Why adaptive monitoring matters
Most monitoring tools rely on fixed thresholds that create noisy alerts and miss real issues.
Instead, we learn the normal behaviour of your accounts and create behavioural baselines, so alert thresholds automatically adapt to campaign changes, seasonality, growth trends, and real-world fluctuations. That means fewer false alarms and faster detection when something genuinely unusual happens.
- You set the number
- Same rule every day of the week
- Doesn't adapt as the account grows
- Fires every Sunday when traffic dips
- Misses gradual drift
- Needs manual maintenance
- Learned from your data
- Knows Sundays are quiet and Mondays spike
- Adapts as spend, traffic and revenue grow
- Only fires on genuine deviations
- Catches gradual drift
- Self-maintaining, unlimited metrics
How often do we check?
Checked as often as feasible, here's a breakdown:
| Platform | How often checked | Alert speed | What we monitor |
|---|---|---|---|
 Google Ads |
Every hour | Within the hour | Spend, CPC, ROAS, conversions, budget pacing |
 Meta Ads |
Every hour | Within the hour | Spend, CPC, ROAS, conversions, budget pacing |
 GA4 |
Daily | Within the day | Sessions, conversions, revenue, engagement |
 Shopify |
Daily | Within the day | Revenue, transactions, AOV, conversion rate |
 Search Console |
Daily | 2–3 day data lag from Google | Organic clicks, impressions, avg. position, keyword clusters |
 LinkedIn Ads |
Daily | Within the day | Spend, clicks, conversions |
Google Ads and Meta data refreshes hourly. We check on every refresh. GA4 and Search Console data is updated by Google once per day. Alerts fire as fast as the data allows.
What types of anomaly do we detect?
Not every anomaly looks the same. A campaign overspending by 200% in a single hour is a different shape from a conversion rate that's been slowly declining for two weeks. We detect both, but they require different detection approaches.
Spike
A sharp single-point change. Spend doubles overnight. Sessions jump 500%. Conversion rate drops to zero. Easy to detect, urgent to act on. Usually a budget edit, a paused campaign, a broken pixel, or a viral social post.
Example: Google Ads daily spend jumps from $200 to $460 in a single day. The baseline expected $210. That's a spike.Drop
The inverse of a spike. A metric falls sharply from its baseline. Drops can be more dangerous than spikes because they go unnoticed. A spike in spend gets attention. A drop in conversions means something broke quietly.
Example: GA4 purchase events go from 23/day to 0. Sessions and spend unchanged. The drop is isolated to conversions, suggesting a tracking issue.Trend shift
A gradual change over days or weeks. No single day looks alarming, but the trajectory is. A campaign drifting from $100 to $140/day over three weeks won't trip a $200 threshold. That's $1,200 of unexpected monthly spend a fixed rule misses entirely.
Example: CPC on a shopping campaign rises from $0.80 to $1.15 over 14 days. No single day is alarming. But the trend is clearly upward and the campaign's profitability is eroding. Go Insights puts this in a watching state and alerts if it continues.Seasonal deviation
A metric breaks from its established weekly pattern. The 90-day baseline learns that Saturdays are always quieter and Mondays always spike. When a Monday performs like a Saturday, it flags the deviation — even though the raw number looks fine in isolation.
Example: GA4 sessions on a Monday come in at 820. The last 12 Mondays averaged 1,400 while Saturdays averaged 780. A fixed threshold set at 500 wouldn't fire, but the baseline knows Mondays should be much higher and flags the 41% shortfall.Absence
A metric that should be there isn't. Zero conversions. Zero spend. Zero clicks. Usually means something is completely broken: a removed pixel, an accidentally paused campaign, a site going down.
Example: Google Ads spend drops to $0 across all campaigns at 3am. No budget changes detected. No campaign pauses. Likely a payment method failure or account suspension.How we catch anomalies before you notice them
Baselines, not thresholds.
Fixed thresholds fire on every Sunday dip and miss gradual drift. Go Insights builds a rolling baseline per metric that learns each account's patterns and only alerts on genuine deviations.
Severity ranking.
Not every anomaly is urgent. Alerts are ranked by severity so you focus on what matters:
Root cause analysis.
Every alert includes a breakdown of related metrics that changed in the same time window. Instead of "conversions dropped 40%," you get "conversions dropped 40%, driven by a mobile conversion rate collapse from 3.8% to 0.3%. Desktop unaffected."
Shopify
How we monitor ad spend and budgets
Daily pacing.
We compare today's spend against expected spend for this time of day, based on the campaign's historical intra-day pattern. Not a flat hourly rate. When pacing drifts, you know within the hour.
Monthly pacing.
We project end-of-month spend based on current daily averages and flag when the projection exceeds or falls short of the monthly budget by more than your threshold.
Budget change detection.
Separate from pacing. When someone edits a budget, pauses a campaign or removes a cap, you're notified immediately. Catches the accidental extra zero, the unintended pause, and the change nobody told you about.
How we detect conversion tracking issues
What we watch.
Conversion count, conversion rate and goal completions across GA4 and Shopify. We alert on sudden drops, gradual declines and unexpected spikes.
How a broken pixel shows up.
A site update removes the conversion tag. Conversions drop to zero, but sessions and spend continue normally. Go Insights flags the gap, and root-cause analysis sees that only conversions changed. That pattern points to a tracking issue, not a performance issue.
How we monitor organic search and keyword rankings
What we watch.
Organic clicks, impressions, average position, and click-through rate from Google Search Console. Keywords are grouped into clusters so you see the health of topics, not individual keyword noise.
Why clusters matter.
A single keyword fluctuating by 2 positions is noise. An entire cluster of 14 related keywords all dropping together is a signal. Clusters smooth out the noise and surface real ranking changes.
Examples of what triggers an alert
- Cluster clicks drop 20%+ week-over-week
- Average position drops 3+ positions across the cluster
- Keywords in top 10 decreases (losing page-one visibility)
- Impressions drop while position is stable (search demand dropped or Google is showing fewer results)
The data lag.
Search Console data arrives 2-3 days late from Google. An alert today reflects what happened 2-3 days ago. SEO changes move slowly, so the lag is usually acceptable.
You choose what's worth hearing about
Each metric has its own customisable sensitivity. Mute what you don't care about, and metrics trending toward a deviation enter a "watching" state, visible on the dashboard but no alert unless it crosses the line.
Loosen sensitivity on noisy metrics. Tighten it on the ones that matter most.
Dismiss what doesn't matter. "Not an issue" teaches the baseline to expect it next time.
What Go Insights detects
Spend pacing anomaly
Google Ads spend increased 63% overnight compared to normal weekday behaviour.
Conversion drop
Lead form submissions dropped while traffic levels remained stable.
Tracking failure
Revenue tracking stopped reporting conversions after a website deployment.
SEO visibility drop
Organic clicks declined sharply outside normal seasonal fluctuations.
Campaign stopped spending
Performance Max campaign unexpectedly stopped spending for 14 hours.
Revenue spike detected
Shopify daily revenue hit a 90-day high driven by organic traffic surge.
Budget set to unlimited
Google Ads campaign budget changed from $200/day to uncapped overnight.
CPA trending up
Cost per acquisition rising steadily over 14 days across non-brand campaigns.
Meta pixel broken
Purchase events dropped to zero while ad spend continued normally.
Keyword cluster improving
Brand keyword cluster moved from position 12 to position 4 over 7 days.
Bounce rate spike
Mobile bounce rate jumped from 45% to 78% after a site update.
Budget decreased 90%
Client reduced daily budget from $500 to $50 without notifying the team.
Spend pacing anomaly
Google Ads spend increased 63% overnight compared to normal weekday behaviour.
Conversion drop
Lead form submissions dropped while traffic levels remained stable.
Tracking failure
Revenue tracking stopped reporting conversions after a website deployment.
SEO visibility drop
Organic clicks declined sharply outside normal seasonal fluctuations.
Campaign stopped spending
Performance Max campaign unexpectedly stopped spending for 14 hours.
Revenue spike detected
Shopify daily revenue hit a 90-day high driven by organic traffic surge.
Budget set to unlimited
Google Ads campaign budget changed from $200/day to uncapped overnight.
CPA trending up
Cost per acquisition rising steadily over 14 days across non-brand campaigns.
Meta pixel broken
Purchase events dropped to zero while ad spend continued normally.
Keyword cluster improving
Brand keyword cluster moved from position 12 to position 4 over 7 days.
Bounce rate spike
Mobile bounce rate jumped from 45% to 78% after a site update.
Budget decreased 90%
Client reduced daily budget from $500 to $50 without notifying the team.
