How to Feed Signed Case Data Back Into Googles AI Bidding Engine

Google's Smart Bidding is one of the most powerful tools in paid search. It uses machine learning to adjust bids in real time, optimizing for the conversion signal you give it. The problem is that most PI firms give it the wrong signal — and then wonder why their Google Ads budget produces a flood of leads but very few signed cases.

The firms seeing 2–3x better ROI from Google Ads aren't using a secret keyword strategy. They're feeding signed-case data back into the bidding engine through offline conversion import, teaching Google's AI to find more people like the ones who actually became cases. This is the highest-ROI technical implementation most PI marketing directors have never done.

How Google's Smart Bidding Actually Works

Smart Bidding strategies — Target CPA, Maximize Conversions, Target ROAS — all rely on the same underlying mechanism. Google's algorithm observes which clicks lead to conversions, identifies the characteristics of those converting users (device, location, time of day, search history, demographics, thousands of signals), and then adjusts bids to find more users who match that profile.

The critical insight: Smart Bidding optimizes for whatever you define as a conversion.It doesn't know what a “good” outcome is for your firm. It only knows what you tell it to count. If you tell it a form submission is a conversion, it will find you more form submissions. If you tell it a phone call over 60 seconds is a conversion, it will find you more 60-second phone calls.

The algorithm is doing exactly what you asked. The question is whether you're asking for the right thing.

The Default Signal Problem: Optimizing for Leads, Not Cases

Most PI firms — and most agencies managing PI accounts — set up Google Ads with standard web conversions: form submissions and phone calls. That's the default because it's easy. Google Tag Manager fires when someone submits a contact form. CallRail or a similar platform reports a call that lasted more than a minute. Done.

But here's what those signals actually measure: interest, not quality. A form submission from someone who was in a fender bender three years ago counts the same as a form submission from someone who was rear-ended last week with documented injuries and $40,000 in medical bills. A 90-second phone call where intake determines the caller has no case counts the same as one that results in a signed retainer.

When Google's algorithm trains on these low-quality signals, it gets very good at finding more of the same. Your lead volume goes up. Your cost per lead might even look reasonable. But your intake team is drowning in unqualified contacts, your conversion rate drops, and your actual cost per signed case keeps climbing. The algorithm is succeeding at the wrong objective.

What Offline Conversion Import Does (And Why It Changes Everything)

Offline conversion import is a Google Ads feature that lets you send conversion data back to Google after the initial click. Instead of telling Google “this click led to a form fill,” you tell it “this click led to a signed case 14 days later.”

Technically, here's what happens: when someone clicks your ad and submits a form or calls, Google records a click ID (called a GCLID). That GCLID follows the lead into your CRM or intake system. When that lead becomes a signed case, you upload the GCLID back to Google Ads with a conversion action labeled “Signed Case” and (optionally) a value.

Now Google's algorithm has a fundamentally different training signal. Instead of learning from thousands of form submissions — most of which never become cases — it learns from the 8–15% that actually signed. It identifies the patterns that distinguish a future signed case from a future rejected lead: the search terms they used, the time of day they searched, the device they were on, their geographic location, their browsing history signals, and hundreds of other factors.

After 60–90 days of receiving this data, the algorithm starts adjusting bids in real time to favor clicks that look like signed cases, not just clicks that look like form submissions. That's the difference between optimizing for volume and optimizing for value.

Step by Step: How to Set Up Signed-Case Feedback

Setting up offline conversion import requires coordination between your Google Ads account, your website, your intake/CRM system, and an attribution layer that connects them. Here's the complete process.

The Data Pipeline: CRM to Attribution to Google Ads

The technical setup above is straightforward in theory. In practice, the challenge is the data pipeline — making sure the GCLID travels cleanly from ad click to CRM to Google Ads without breaking.

Where this pipeline commonly breaks:

• GCLID not captured on the form.If your website doesn't have a hidden field pulling the GCLID from the URL, the chain breaks at step one. This is a 15-minute fix for any developer, but it's the most common gap.
• CRM doesn't have a GCLID field.Many PI firms never set up a custom field for the GCLID because nobody told them they'd need it. If the GCLID isn't stored, you can't upload it later.
• No automation for the upload.Manually downloading a CSV from your CRM and uploading it to Google Ads every week is technically possible, but it doesn't happen consistently. The upload needs to be automated or it will be abandoned within a month.
• Conversion window mismatch. If your Google Ads conversion window is set to 30 days but your average time from click to signed case is 45 days, Google will never see those conversions. Set the window to 90 days to account for PI intake timelines.

A revenue intelligence platform that integrates with both your CRM and Google Ads can automate this entire pipeline — capturing the GCLID, tracking the lead through intake, and pushing the signed-case conversion back to Google without manual intervention.

Expected Impact: What Firms See After 60–90 Days

The algorithm needs data to learn. Expect 60–90 days of “training” before you see meaningful changes in bidding behavior. During this period, Google is collecting signed-case conversions and building a model of what a high-value click looks like for your firm.

After the training period, the changes are measurable and consistent:

The counterintuitive result: your cost per lead often goes up after implementing offline conversions. Google is bidding more aggressively on the clicks that are likely to become cases, and those clicks are more competitive. But your cost per signed case drops significantly because a much higher percentage of those leads convert.

For a firm spending $50,000/month on Google Ads, cutting cost per signed case from $4,000 to $2,000 means doubling the number of signed cases from the same budget. That's not an incremental improvement. That's a structural change in your marketing economics.

Why Most PI Firms Haven't Done This

If the ROI is this clear, why aren't more firms doing it? Three reasons, all related:

Data silos. The ad click data lives in Google Ads. The lead data lives in your CRM or intake system. The case outcome data lives in your case management software. Nobody in the firm has visibility across all three systems, so nobody is connecting the dots. The marketing director sees leads. The intake manager sees conversions. The managing partner sees case outcomes. But no single person or system tracks a lead from ad click to signed case.

Technical complexity.Setting up GCLID capture, CRM storage, and automated uploads requires coordination between your web developer, your CRM administrator, your marketing team, and your Google Ads manager (or agency). In most PI firms, these are different people (or different vendors) who don't have a shared project driving this forward. The marketing director would need to manage a cross-functional implementation that touches four different systems, and that project keeps getting deprioritized.

Nobody is asking the right question.Most agencies report on cost per lead and lead volume because those are the metrics they can measure. They don't ask “what happened to those leads after they hit your CRM?” because the answer requires data they don't have access to. And most marketing directors don't know that feeding case data back into Google is even an option — it's not something Google promotes heavily because it requires work outside their ecosystem.

The result is that the vast majority of PI firms are leaving 2–3x ROI on the table from their existing Google Ads spend. Not because the platform doesn't work, but because they're giving it the wrong instructions.

The Bottom Line

Google's Smart Bidding is only as good as the conversion signal you feed it. If you're telling Google to optimize for form submissions, you'll get more form submissions — most of which will never become cases. If you tell Google to optimize for signed cases, you'll get fewer but dramatically better leads, and your cost per signed case will drop.

The technical setup is manageable. The data pipeline is solvable. The hard part is connecting the systems and maintaining the flow of case data back to Google on a consistent basis. That's where a marketing attribution platform built for PI firms becomes the bridge between your CRM and your ad accounts — automating the closed loop so Google's AI works for your firm, not just for your lead count.

If you're spending $30,000 or more per month on Google Ads and you're not feeding case data back into the bidding engine, this is the single highest-ROI project your marketing team can take on this quarter. Forward this to your agency and ask them if they've set it up. If they haven't, now you both know what to do.