Google AdsSmart BiddingOffline Conversion TrackingLTV Optimization

How We Drove a 73% ROAS Increase for Gusto with Offline Conversion Tracking & Smart Bidding

For Gusto's 2020 Google Ads program, TNT Growth integrated Gusto's lifetime value model directly into Smart Bidding through offline conversion tracking — turning raw LTV data into a 73% non-brand ROAS increase and a 127% lift in conversion rate.

Gusto · 2020 Campaign · HR & Payroll SaaS · San Francisco, CA

73%Non-Brand ROAS Increase

+21%Branded ROAS Increase

127%Conversion Rate Lift

45%ROAS Improvement (6-Wk Test)

6 WksTo Prove the Methodology

Gusto: A Modern HR Platform With a High-Value Conversion Problem

Gusto is the go-to HR platform for modern small businesses — payroll, benefits, 401(k), expert HR, and team management in one place. When you're competing for high-LTV customers in a market full of well-funded competitors, paid search isn't just a traffic channel — it's a revenue engine that needs to be calibrated to business outcomes, not just click volume.

Gusto had already done the hard work of building a sophisticated lifetime value model — a proprietary signal that distinguished their most valuable customers from lower-value cohorts. The problem: that signal was locked inside Gusto's internal systems and wasn't influencing how their Google Ads campaigns bid. They had integrated their LTV model into a third-party automated bidding solution, but the results weren't moving meaningfully.

The mandate was clear: unlock the full value of Gusto's LTV data by integrating it directly with Google's machine learning through Smart Bidding. The hypothesis was that Google's native bidding algorithms — fed accurate, revenue-weighted conversion signals via offline conversion tracking — would dramatically outperform any third-party solution that couldn't access Google's full data infrastructure.

The Mandate

73%

ROAS increase. On non-brand. After full migration.

"Stop optimizing for conversion volume. Start optimizing for customer lifetime value."

Three Structural Problems Preventing Gusto's LTV Data From Driving Campaign Performance

Attribution Without LTV Context

Every conversion in Google Ads was treated equally — a trial signup from a solo freelancer and a signup from a 50-person company with high renewal probability looked identical to the bidding algorithm. Without LTV-weighted signals, Smart Bidding couldn't distinguish between low-value and high-value customers, so it optimized for volume rather than revenue.

Third-Party Bidding Limitations

Gusto's existing setup piped LTV data into a third-party automated bidding tool, but the results weren't materializing. The fundamental constraint: third-party bidding solutions operate outside Google's core machine learning infrastructure. They can't access the same real-time auction signals, Quality Score data, and user intent models that Google's native Smart Bidding algorithms process at the moment of each auction.

Untapped First-Party Data

Gusto possessed something most advertisers don't: a sophisticated, validated LTV model built on years of customer data. That signal — if properly imported into Google Ads — could teach the algorithm to bid higher on searches likely to produce high-LTV customers and lower on searches likely to produce churn risk. The challenge was building the technical pipeline to get that data into Google Ads in real time.

Building the Attribution Foundation

The foundational move was implementing offline conversion tracking (OCT) — a Google Ads feature that allows businesses to import conversion data from outside the ad platform, including CRM data, revenue data, and predicted LTV scores. For Gusto, this meant building a pipeline that passed LTV-weighted conversion values back into Google Ads, so every recorded conversion carried a value signal reflecting the projected lifetime revenue of that customer, not just a binary conversion event.

Once OCT was live, the algorithm's understanding of a "successful conversion" fundamentally changed. Instead of treating every signup as equivalent, Smart Bidding could now see that certain search queries, audiences, and creative combinations produced customers worth 3X, 5X, or 10X the lifetime revenue of others. That signal became the raw material for everything that followed.

Replacing Third-Party Bidding With Google's Native Machine Learning

With accurate LTV-weighted conversion signals flowing into Google Ads, the team migrated campaigns from the third-party bidding solution to Google Smart Bidding — specifically Target ROAS bidding, configured with conversion values that reflected actual customer lifetime value rather than uniform conversion counts. This was the unlock: Google's Smart Bidding, fed by accurate value signals, could now do what the third-party tool could not.

The key advantage of native Smart Bidding is its access to Google's real-time auction data — Quality Score signals, device, location, time-of-day, audience overlap, search query context — factors that third-party tools can only approximate. When you combine that real-time signal richness with accurately weighted conversion values, the algorithm has everything it needs to make high-quality bid decisions at scale.

Testing the Methodology Before Committing

127%Conversion rate increase during the 6-week Drafts & Experiments test period

Before migrating the full account, TNT Growth used Google Ads' Drafts & Experiments feature to run a controlled A/B test — splitting budgets on select campaigns into control groups (existing third-party bidding) and test groups (Smart Bidding with LTV-weighted OCT signals). This de-risked the transition and created a clean comparison: same campaigns, same budget split, different bidding methodology.

The 6-week test period delivered a 127% increase in conversion rate and a 45% improvement to ROAS in the test groups — a statistically meaningful result that validated the LTV+Smart Bidding thesis before a full account migration. When the data confirmed what the hypothesis predicted, the migration decision was easy.

Scaling What the Test Proved

With the experiment validating the approach, Gusto migrated all campaigns to Smart Bidding. Non-brand campaigns — where new customer acquisition happens and LTV differentiation matters most — delivered a 73% ROAS increase. Branded campaigns, where search intent is already highly qualified, saw a +21% ROAS improvement as Smart Bidding optimized bid levels against revenue-weighted signals rather than flat conversion counts.

Four Strategic Decisions That Drove Outsized ROAS Growth

LTV as a Bidding Signal

Passing Gusto's LTV model into Google Ads via OCT transformed bidding from "optimize for conversions" to "optimize for revenue." The algorithm learned to recognize the search behaviors, audience attributes, and creative combinations that predicted high-LTV customers — and bid accordingly, shifting spend toward higher-quality acquisition opportunities.

Controlled Experiment First

Using Drafts & Experiments before committing to a full migration was the move that de-risked the entire strategy. The 6-week test produced a 127% CVR lift and 45% ROAS improvement under controlled conditions — removing the uncertainty that typically accompanies major bidding architecture changes.

Non-Brand Campaign Prioritization

Non-brand campaigns are where LTV differentiation creates the most leverage: these are users in active consideration mode, not already brand-loyal. Smart Bidding with LTV signals systematically shifted spend toward non-brand queries that produced high-value customers, explaining why the 73% ROAS improvement was concentrated in this campaign type.

Native Over Third-Party

The core strategic insight was that Google's Smart Bidding, fed accurate LTV signals, outperforms any third-party bidding solution — because third-party tools can't access Google's real-time auction intelligence. The LTV data was always there. The missing piece was piping it directly into the infrastructure that could use it most effectively.

The Results: 73% ROAS Growth Driven by LTV-Weighted Smart Bidding

The results validated the core hypothesis: Google's Smart Bidding, fed accurate lifetime value signals through offline conversion tracking, dramatically outperforms third-party bidding solutions. Here's what the full migration delivered for Gusto:

73%Non-Brand ROAS IncreaseAfter full Smart Bidding migration — the primary new-customer acquisition channel.

+21%Branded ROAS IncreaseSmart Bidding with LTV signals improved performance even in already high-intent brand campaigns.

127%Conversion Rate LiftAchieved during the 6-week Drafts & Experiments test period before full migration.

45%ROAS Improvement (Test)Test groups outperformed control groups by 45% in ROAS during the controlled experiment phase.

What Made This Work: Treating LTV Data as a Bidding Input, Not Just an Analytics Metric

The Gusto campaign succeeded because TNT Growth reframed what LTV data is for. Most advertisers use LTV data to understand cohorts in retrospect — to analyze which customers were most valuable after the fact. The strategic leap was treating LTV as a real-time bidding input: a signal that could be fed into Google's machine learning at the moment of every auction to shift spend toward acquisition opportunities that predicted high-lifetime-value customers.

The Drafts & Experiments methodology was equally critical. Migrating a full account from a third-party bidding solution to Smart Bidding without validation carries real risk — the algorithm needs time to learn, and a poor transition can damage performance during the learning period. By testing in a controlled environment first and validating a 127% CVR lift before committing, the team removed that risk and entered the full migration with confidence.

"The 45% ROAS improvement in the test wasn't just a good number — it was proof that LTV-weighted Smart Bidding was the right architecture. Everything after that was execution."

— Adam Treboutat, Growth Marketing Lead, Gusto

Ready to Scale?

Stop Leaving Revenue on the Table

Book a free strategy call with our team. We'll show you exactly where the profit is hiding in your current ad spend.

20% Lift Guarantee

We guarantee a 20% lift in conversions within 90 days — or it's free. That's how confident we are in our system.

Schedule a Free Call

No commitment. No pressure. Just a real conversation about growth.