The 7 Anomalies Every PI Firm's Alert System Should Catch Automatically

PI marketing portfolios generate dozens of data points every day — lead volumes, costs, conversion rates, contact rates, signing pace. Most of that data falls within normal operating ranges. But buried in the noise are anomalies that signal real problems: a vendor quietly underdelivering, a cost metric drifting upward, a conversion funnel breaking down at a specific stage.

The firms that catch these anomalies early save $10,000–$30,000 per incident. The firms that catch them in monthly reviews — or never — absorb the full cost. Here are the seven anomaly types that cause the most financial damage in PI marketing, and exactly what your alert system should look for in each case.

1. Vendor Delivery Drop

What It Looks Like

A vendor that normally delivers 40–50 leads per week suddenly drops to 25–30. The decline may happen all at once or gradually over 7–10 days. Because daily lead counts are naturally volatile, the pattern often hides until you look at weekly rolling averages.

Why It Matters

When a vendor reduces delivery without reducing your invoice, your effective CPL increases even though the stated CPL stays the same. A vendor billing $10,000/month for 50 leads ($200 CPL) that drops to 30 leads is now costing you $333 per lead. If you don't catch it for a month, you've overpaid by roughly $4,000.

Alert Configuration

Monitor 7-day rolling lead volume by vendor. Trigger an informational alert at a 20% drop from the 90-day weekly average. Trigger a warning at 30%. Trigger a critical alert at 50% or if the drop persists for 10+ consecutive days. Use a minimum 3-day duration filter to avoid flagging normal weekly variability.

2. Cost-Per-Lead Spike

What It Looks Like

A vendor's CPL jumps from $180 to $240+ and stays elevated. This can result from the vendor changing their traffic mix, losing a high-performing campaign, or shifting your budget to a more expensive channel without telling you.

Why It Matters

A 30% CPL spike on a $15,000/month vendor costs you $4,500 in extra spend per month — or $1,125 per week. At that rate, every week you don't catch the spike is another $1,000+ burned. Over 30 days, the total waste typically lands between $15,000 and $25,000 depending on vendor spend level. For a detailed walkthrough of this exact scenario, see what happens when a PI firm ignores a CPL spike for 30 days.

Alert Configuration

Calculate a 90-day rolling CPL average per vendor. Informational alert at 15–20% above baseline. Warning at 25–35%. Critical at 40%+. Require 5-day sustained elevation before triggering warning or critical to filter out single-invoice timing anomalies.

3. Conversion Rate Collapse

What It Looks Like

A vendor's lead-to-signed-case conversion rate drops from 10% to 6% or lower. This is the most insidious anomaly because the vendor's CPL and lead volume may look completely normal. The problem only appears when you track what happens after the lead arrives.

Why It Matters

Conversion rate collapses directly increase your cost per signed case. A vendor with a $200 CPL and 10% conversion costs $2,000 per signed case. At 6% conversion, the same $200 CPL vendor costs $3,333 per signed case — a 67% increase that won't show up in any vendor-provided report.

Alert Configuration

Use a 90-day rolling conversion rate by vendor as your baseline. This metric needs a longer observation window because conversion data lags lead delivery by days or weeks. Informational at 20% below baseline. Warning at 30%. Critical at 40%+. Use a 7-day minimum duration filter — conversion data is noisier than volume or cost data.

4. Budget Overspend Pace

What It Looks Like

By day 12 of the month, a vendor has consumed 55% of their monthly budget instead of the expected 40%. If the pace continues, they'll exceed budget by 20–30% by month end.

Why It Matters

Budget overspend is the most directly controllable anomaly on this list. Unlike conversion rate shifts, which require investigation, overspend pace requires a single action: reduce daily caps or pause the vendor. A $20,000/month vendor on pace to spend $26,000 wastes $6,000 unless you intervene. That's $6,000 that could fund a higher-performing vendor or be held as profit.

Alert Configuration

Calculate expected daily spend (monthly budget ÷ days in month). Track cumulative spend vs. expected cumulative spend. Warning when actual exceeds expected by 15%+ at the day-10 mark. Critical when it exceeds expected by 20%+ at any point after day 7. No duration filter needed — this is a running calculation, not a point-in-time metric.

5. Intake Contact Rate Drop

What It Looks Like

Your intake team normally contacts 80–90% of new leads within 5 minutes. That rate drops to 60–70%. The leads are arriving, but your team isn't reaching them fast enough.

Why It Matters

Speed-to-contact is the single biggest controllable factor in lead-to-case conversion. Studies consistently show that contacting a PI lead within 5 minutes versus 30 minutes can double the conversion rate. A 20% drop in contact rate typically produces a 10–15% drop in signed cases from the same lead volume — and it's entirely a staffing or process problem, not a vendor problem.

Alert Configuration

Monitor daily contact-within-5-minutes rate. Informational at 10% below target. Warning at 20% below target. Critical at 30%+ below target or if contact rate drops below 65% for any 48-hour period. This alert should route to the intake manager, not the marketing director.

6. Case Signing Pace Decline

What It Looks Like

Your firm signs 25–30 cases per week from paid sources. Over two weeks, that drops to 18–22 without an obvious cause. Lead volume looks stable. Contact rates look normal. But fewer leads are converting to signed cases.

Why It Matters

A signing pace decline is often the downstream effect of an upstream anomaly that went undetected — a vendor quality shift, a conversion process change, or a criteria adjustment. Each unsigned case represents $3,000–$5,000 in acquisition cost that produced nothing, plus the lost settlement revenue. At 5 fewer signings per week, the financial impact reaches $15,000–$25,000 per week in wasted spend alone.

Alert Configuration

Track 7-day rolling signed case count. Informational at 15% below the 90-day weekly average. Warning at 25%. Critical at 35%+. This is a lagging indicator, so pair it with the leading indicators above (conversion rate, contact rate) for faster detection.

7. Settlement Value Shift

What It Looks Like

The average settlement value for cases from a specific vendor drops from $45,000 to $32,000 over a 6-month window. The vendor is still delivering leads that convert to signed cases, but the cases are worth less.

Why It Matters

This is the longest-horizon anomaly and the hardest to detect without revenue intelligence. A vendor with a $2,500 cost per case and $45,000 average settlement delivers 18x ROI. The same vendor at $32,000 average settlement delivers 12.8x ROI — still positive, but a 29% decline that fundamentally changes how you should allocate budget. If you track only cost per lead or cost per case, you miss this entirely.

Alert Configuration

This metric requires a 6–12 month lookback due to settlement lag in PI cases. Monitor quarterly rolling average settlement value by lead source. Informational at 10% below the trailing 4-quarter average. Warning at 20%. Critical at 30%+. Review quarterly rather than weekly — this metric doesn't move fast enough for real-time alerts.

Building a Complete Detection System

Each of these seven anomalies represents a different failure mode in your marketing portfolio. Some are fast-moving (CPL spikes, volume drops) and need real-time monitoring. Others are slow-moving (settlement value shifts, signing pace declines) and need weekly or monthly review. A complete system monitors all seven at the appropriate cadence.

For a step-by-step guide to configuring thresholds and routing for each anomaly type, read How to Configure Performance Alerts That Catch Problems Before They Cost You Money. And to see how AI-powered anomaly detection handles all seven categories automatically, including the baseline calculations, threshold calibration, and alert routing that would take hours to maintain manually — that's exactly the problem RevenueScale was built to solve.

The difference between firms that waste $50,000–$100,000 per year on undetected marketing anomalies and firms that catch problems in hours isn't talent or attention. It's systems. Build the system, and the savings follow.