Search lead quality collapse after broad-match expansion
A B2B workflow-software company spends $18,000 per month on Search across three campaigns: Brand, Non-brand core, and Competitor. Over eight weeks, paid leads rose 35%, but the sales-qualified rate fell from 22% to 9%. Non-brand uses broad match with Maximize conversions; Brand receives 28% of spend and 41% of reported conversions. Search impression share is 54%, with 19% lost to budget. Last month the team added a 'free demo' landing page and a call extension; calls shorter than 60 seconds count as conversions. CRM records show 38% of paid leads lack company-size data, and sales says many leads are students or sole proprietors. Reported cost per lead is $74, below the $95 target. The VP wants 25% more volume next quarter and asks whether to raise budget or expand broad match.
1. What should the team validate before approving the 25% budget increase or broader match? Explain the evidence you would use.
Read the worked answer
Do not approve the increase yet. The headline lead volume is misleading. If prior monthly leads were L, the old qualified volume was 0.22L; the new volume is 1.35L × 0.09 = 0.1215L, roughly a 45% decline in sales-qualified leads. The reported CPL of $74 therefore hides a cost per qualified lead near $822, far above what the $95 lead target implies.
Validate three things before spending more:
- Conversion definition: calls under 60 seconds and 'free demo' requests are low-intent actions. Segment conversion rate and qualified rate by conversion action, campaign, query, device, and landing page.
- Lead quality data: 38% missing company size means qualification is incomplete. Match CRM records to Ads leads and compare qualified rate for broad-match queries versus phrase/exact and brand.
- Brand distortion: Brand is 28% of spend but 41% of conversions. That may be efficient, but it can make blended CPL look healthy while non-brand quality deteriorates. Check brand incrementality separately.
The 19% budget-lost impression share shows demand exists, but adding budget to a Maximize conversions campaign optimizing raw leads would buy more of the wrong signal. Expanding broad match before fixing qualification would amplify the problem. First prove which segments produce SQLs, then decide whether incremental budget has a qualified-lead case.
2. Recommend changes to conversion tracking, campaign structure, and bidding. Prioritize them and explain the trade-offs.
Read the worked answer
Priority 1 - change what the account optimizes toward. Create a qualified-lead conversion from CRM stage or from form fields that require company size and role. Import offline SQL and closed-won stages using GCLID where available and enhanced conversions for leads with hashed email. Keep raw leads as a diagnostic, not the primary bid signal. Change the call conversion to a qualified-call threshold, such as 90–120 seconds, or import calls that CRM marks as meaningful.
Priority 2 - separate intent and control brand. Keep Brand in its own exact/phrase campaign so non-brand budget cannot suppress it, but measure brand incrementality rather than assuming all brand conversions are incremental. Split Non-brand into core solution and competitor, with competitor budget capped. Use phrase/exact for proven terms and broad only in a controlled campaign with search themes or audience signals.
Priority 3 - bid on value. Once enough offline qualified conversions exist, use tCPA on qualified leads or Maximize conversion value with differentiated values for SQL and closed-won. Use portfolio bidding only if campaigns share a comparable value definition.
Add negatives for free, student, template, and job-seeker queries; use landing-page qualifiers and form routing rather than trying to negative-match job titles. Expanding broad match is reasonable only after negatives, conversion quality, and offline feedback are in place.
3. How would you prove whether added budget creates incremental qualified pipeline rather than pulling forward demand?
Read the worked answer
Use an incrementality design, not a before-and-after report. A geo holdout or conversion lift test is appropriate: match similar regions or accounts, randomize which receive the extra budget, and compare qualified pipeline and closed-won revenue over a pre-registered period. Include a long enough window to cover the 30–90-day sales cycle and avoid measuring only immediate form fills.
For brand, run a separate holdout because brand search often captures demand that would have converted anyway. For non-brand, compare cost per incremental SQL, not cost per click or reported lead. Keep a control group and account for seasonality, sales capacity, and concurrent launches.
Operationally, use the test to answer a narrow question: does a 25% budget increase produce at least a 25% increase in qualified pipeline at acceptable unit economics? If the result is flat, the constraint is lead quality or sales follow-up, not budget. Do not use last-click ROAS from the same campaign as proof of incrementality.