How to Set Up Google Ads Offline Conversion Import for B2B Lead Gen

Google's Smart Bidding learns from your conversion data. The system observes which clicks produced conversions, identifies patterns in the users and queries that generated those conversions, and adjusts bids to find more of the same.

The problem for B2B advertisers: most Google Ads accounts are sending form fills as the primary conversion signal. Smart Bidding learns to find people who fill out forms. Those are not the same as people who buy.

Offline Conversion Import (OCI) closes that gap. It lets you pass downstream CRM events back to Google so the system optimizes toward outcomes that connect to revenue: qualified opportunities, accepted deals, or whatever stage in your pipeline predicts closed-won most reliably. Here is how the mechanic works and how to set it up.

Why Form Fill Conversions Are the Wrong Signal

When a visitor clicks your Google ad and submits a form, Google records a conversion. The system notes everything about that click: the search query, the device, the audience segments the user belonged to, the keyword match type, the time of day. It uses all of that to build a model of what a converting user looks like.

The model gets accurate at its goal. The problem is the goal is wrong.

Form fill conversion rates are noisy for B2B accounts because the form is the first filter, not the last one. Submitters include vendors, students, consultants, ICP-fit contacts with no budget authority, and genuine in-market buyers. Smart Bidding treats them identically because from Google's perspective, they all converted.

If form fills are your only conversion event, you are asking the system to optimize for a proxy metric. It will get very good at producing form fills. Your sales team will continue to flag lead quality problems. The disconnect is not in your targeting or your creative. It is in the signal you are feeding the bidding system.

How Offline Conversion Import Works

The mechanic has three steps.

Step 1: Capture the gclid. When a visitor clicks your Google ad, Google appends a unique click identifier to the landing page URL. This is the gclid (Google Click Identifier). Your landing page needs to capture that value and pass it through the form submission into your CRM. This typically requires a small JavaScript snippet that reads the gclid from the URL parameters and populates a hidden form field. Most CRMs (Salesforce, HubSpot, Marketo) have native gclid capture features or documented implementation guides. Verify it is actually working by checking that recent Google-sourced contacts in your CRM have a gclid value populated.

Step 2: Store the gclid on every Google-sourced CRM record. When a contact progresses through your pipeline, the gclid travels with the record. When that contact becomes a qualified opportunity or a closed deal, your CRM can export that event with the associated gclid and the date it occurred.

Step 3: Upload the conversion data to Google. Google accepts offline conversion data via CSV upload in the Google Ads interface, via the Google Ads API, or through native integrations with tools like HubSpot, Salesforce, and Google Ads Data Manager. Each row in the upload includes the gclid, the conversion action name, the conversion time, and optionally a conversion value. Google matches the gclid back to the original click and credits the correct campaign, ad group, keyword, and creative.

Enhanced Conversions for Leads: The Lower-Overhead Alternative

If gclid capture is not yet implemented in your setup, Enhanced Conversions for Leads offers a path to downstream signal without the full gclid infrastructure.

Instead of matching on gclid, Enhanced Conversions for Leads matches on hashed first-party data: the email address or phone number collected in the form submission. When you upload a conversion event with a hashed email, Google looks up whether that email corresponds to a signed-in Google user and matches it back to an ad interaction from that user.

Match rates are lower than gclid-based OCI, typically 40 to 60 percent for B2B accounts where work email addresses often differ from personal Google accounts. Gclid-based matching runs 80 to 90 percent when capture is implemented correctly.

Enhanced Conversions for Leads is the right starting point when gclid capture needs development work. It is not a full replacement for gclid-based OCI when you need the strongest signal fidelity for Smart Bidding.

What Conversion Event to Import

The conversion event you import is the most consequential decision in the OCI setup.

Form fill is already tracked as a website conversion action. Do not also import it via OCI unless you are using it as an observation-only micro-conversion. Importing the same signal twice inflates conversion volume and distorts Smart Bidding targets.

Marketing Qualified Lead (MQL) is better than form fill but still carries noise for reasons covered in the MQL signal migration post. If MQL is the highest-fidelity downstream event you can access, import it as a starting point. Understand that Smart Bidding will optimize toward MQL-quality submissions, and your MQL definition may include contacts that never produce pipeline.

Qualified opportunity is the recommended primary conversion event for most B2B Google Ads accounts. A qualified opportunity means a human reviewed the lead, confirmed it as ICP-fit and in active evaluation, and opened a sales cycle. This is the signal that correlates most directly with revenue. Import opportunity creation as your primary OCI conversion action and set your Smart Bidding tCPA target against cost per opportunity, not cost per lead.

Closed-won is the cleanest signal, but the lowest volume. For accounts generating fewer than 20 to 30 closed deals per month from Google, closed-won data is too sparse for Smart Bidding to use as a primary signal. Use it as a secondary conversion action with observation-only status while your opportunity volume builds. As closed-won volume grows, you can layer it in as a weighted signal.

The Conversion Lag Problem

B2B sales cycles are long. A click that generates a qualified opportunity might have happened 30, 60, or 90 days before the opportunity was created in your CRM. Google's OCI system accepts conversions up to 90 days after the originating click, which covers most B2B sales cycles for the opportunity creation event.

The implication for upload cadence: when you upload opportunity data this week, some of those conversions will reference clicks from 6 or 8 weeks ago. That is expected and correct. The gclid stays valid for 90 days from the original click.

For accounts where the average time from click to opportunity exceeds 90 days, closed-won attribution will fall outside the matching window for many deals. Opportunity creation (which happens earlier in the cycle) is a more reliably matchable event in those cases.

One setting that most B2B teams miss: the conversion window in Google Ads under each conversion action's settings. The default is often 30 days. At 30 days, any conversion that happens more than 30 days after the click is not attributed. Change this to 90 days for all OCI conversion actions. Leaving it at the default systematically understates the value of your campaigns and degrades the Smart Bidding signal.

Upload Cadence

Weekly uploads are the standard. OCI does not require real-time data. It requires data fresh enough that Smart Bidding can act on recent signal without waiting months to update its model.

For accounts with low opportunity volume (fewer than 10 per month from Google), bi-weekly uploads are acceptable. For accounts with Salesforce or HubSpot integrated directly into Google Ads Data Manager, daily automated uploads are the cleanest setup. The integration handles the export and upload without manual intervention.

Avoid large historical batch uploads covering multiple months at once. Uploading 6 months of past conversions in one file can cause Smart Bidding to overfit to patterns from campaigns and targeting configurations that no longer reflect how your account is set up.

What This Changes for Smart Bidding

Once opportunity data starts flowing, the tCPA target you set in Smart Bidding needs to change.

The cost per opportunity will be significantly higher than your cost per form fill, because you are measuring a different and harder-to-reach outcome. A B2B account that was running tCPA at $80 per form fill might set tCPA at $600 to $1,200 per qualified opportunity. Both numbers can be right for the same account. The difference is what Google is being asked to optimize toward.

Recalculate your acceptable cost per opportunity based on your pipeline conversion rates and deal size. If your average deal is $40,000, your close rate from opportunity is 25 percent, and your target CAC is $8,000, you can afford up to $2,000 per qualified opportunity from paid search. That is your tCPA ceiling.

Setting tCPA based on what sounds reasonable rather than what your pipeline math supports is one of the most common reasons OCI implementations underperform. The math should set the target, not an intuition about what a lead should cost.

What to Test

Two steps.

Step 1: Audit your current conversion setup in Google Ads. Under Tools and Settings, check which conversion actions are set to "Include in conversions." If form fills are your only primary conversion action, you are at the most common misconfiguration in B2B Google Ads accounts. Verify gclid capture is working on your most recent Google-sourced CRM contacts.

Step 2: Once opportunity uploads begin, create a new Smart Bidding experiment in your highest-spend campaign. Set the experiment bid strategy to target CPA at your calculated cost-per-opportunity ceiling. Let the experiment run for at least 4 to 6 weeks to allow Smart Bidding to build a model on the new signal. Compare cost per opportunity, pipeline value, and CAC between the control and experiment. If the opportunity-optimized campaign produces better pipeline per dollar, roll the bid strategy change to the full account.

Accounts that import opportunity or closed-won data consistently outperform accounts optimizing on form fills at the Smart Bidding layer. The quality advantage compounds over time as the model builds a more accurate picture of what an in-market buyer looks like for your specific account and product.

What's changing in B2B paid media. What it means for your pipeline.

Subscribe free →