The Challenge

Complex civil litigation is not an impulse purchase. A business owner or high-net-worth individual facing a serious civil case does not click one Google ad and call immediately. They research extensively, visit multiple attorney websites, read reviews, come back days later, and make a decision over a consideration cycle that can stretch weeks.

The measurement problem this creates is severe. Standard last-click attribution - the default in most Google Ads accounts - gives 100% of the conversion credit to the final keyword a prospect clicked before submitting a contact form. Every prior touchpoint across that 7 to 12 step journey gets zero credit. The keywords and ads that built awareness, established authority, and drove early research - invisible.

  • Last-click attribution hiding the true value of early-funnel and mid-funnel keywords
  • Budget being cut from awareness and consideration keywords that were actually driving pipeline
  • Cross-device journeys breaking attribution entirely - desktop research, mobile form submission
  • Offline call conversions not being matched back to original ad clicks
  • Smart Bidding algorithm optimizing on incomplete and misleading conversion data

Rebuilding the Attribution Model

The first move was eliminating last-click attribution entirely and replacing it with a custom position-based weighting model. Position-based attribution distributes credit across the full touchpoint journey rather than awarding it all to the final click.

40/20/40 Position-Based Attribution Model
First touchpoint: 40% of conversion credit (awareness entry point)
Middle touchpoints: 20% distributed evenly across consideration journey
Final touchpoint: 40% of conversion credit (decision trigger)

Rationale: For high-consideration legal services, both the entry keyword that initiated research AND the decision keyword that closed the inquiry deserve significant weight. The mid-funnel consideration keywords share the remaining 20%.

This model immediately changed which keywords the account was optimizing for. Broad informational queries like "what to do if sued by a business" and "commercial litigation process" - previously showing zero conversion value under last-click - began receiving partial attribution credit as first-touch entry points. Budget that had been cut from these terms was restored, and the algorithm began bidding for the full funnel rather than just the bottom of it.

Enhanced Conversions and Cross-Device Recovery

Attribution model changes alone do not solve the cross-device problem. A prospect who starts researching on a work desktop and submits a contact form from a personal phone is treated as two completely separate users under standard cookie-based tracking. The desktop research session - potentially containing the highest-value keyword clicks - is orphaned.

SHA-256 Enhanced Conversions setup: When a prospect submits the contact form, their email address is captured, hashed using SHA-256 encryption on the server side, and passed directly to Google Ads via the Enhanced Conversions API. Google matches the hashed email against logged-in Google account data to stitch cross-device journeys together - connecting the desktop research session to the mobile form submission within the same conversion path. No PII is ever transmitted in plain text.

The cross-device recovery rate from Enhanced Conversions was material. A significant portion of conversions that had been appearing as direct or unattributed traffic were revealed to be Google Ads touchpoints from earlier in the same prospect journey - just on a different device.

Offline Call Matching

For a complex litigation firm, a meaningful portion of conversions happen over the phone rather than through online forms. These calls were previously invisible to the Google Ads attribution system - meaning every prospect who called after clicking an ad was contributing zero data back to the algorithm's optimization logic.

  • Dynamic call tracking numbers deployed per campaign to isolate call source at the keyword level
  • Call recordings reviewed by intake team and qualified calls marked in the CRM
  • Qualified call events fired as offline conversions back to Google via webhook postback within 24 hours
  • Call duration threshold set at 4 minutes minimum before a call counted as a qualified conversion event
  • Unqualified calls explicitly logged as negative signals to inform audience exclusions

Why call duration matters as a signal: In complex litigation, a 90-second call is almost always someone asking a basic question or calling the wrong number. A 4-minute-plus call indicates the prospect shared enough detail about their situation to warrant a real intake conversation. Setting the threshold at 4 minutes meant only genuinely qualified prospect calls fed back into the Google bidding algorithm - not misdials or irrelevant inquiries. This is why average intake call duration increasing 62% was a meaningful outcome: the algorithm learned to find more of the right callers.

Results

The attribution rebuild took approximately 45 days to accumulate enough new conversion data to meaningfully shift the Smart Bidding algorithm's behavior. From that point, the compounding effect was significant.

  • 28% more profitable keyword data recovered from cross-device journey stitching and position-based attribution
  • Average intake call duration increased 62% as the algorithm optimized toward higher-intent searchers
  • Early and mid-funnel awareness keywords restored to budget allocation based on their true attribution value
  • Google Smart Bidding performance improved as conversion signal quality compounded over subsequent months
  • Offline call conversions now feeding the algorithm directly for the first time since the account launched

Client name withheld per confidentiality agreement. Results measured over a 90-day period following attribution model implementation.