A B2B SaaS company optimized Google Ads for closed revenue instead of form fills and scaled pipeline by 80% in 90 days. This is a case study in what happens when you stop letting Google’s bidding algorithms chase vanity conversions and start feeding them the signals that actually matter to your business. The company in question is a mid-market B2B software firm with a two-person marketing team, roughly $40K per month in Google Ads spend, and an 18-month relationship with a traditional agency. On paper, everything looked healthy. Cost per lead was declining, volume was climbing, and the agency’s reports were green across the board. The problem was that almost none of those leads were turning into revenue. This is the story of how they diagnosed the root cause, rebuilt the signal stack, restructured every campaign, and came out the other side with a pipeline that actually converted.
The Situation: Leads Were Up, Revenue Was Flat
The Business And The Setup
The company sells workforce management software to mid-size companies, with average contract values in the $30K to $50K range and a sales cycle that typically runs 60 to 90 days from first touch to closed deal. Their marketing team consisted of a head of demand gen and a marketing coordinator. Google Ads was the primary paid acquisition channel, responsible for feeding the top of a pipeline that the sales team worked through Salesforce.
Their agency had been managing the account since launch. The relationship was structured around a percentage-of-spend fee with a six-month contract, and the agency delivered monthly reports focused on impressions, clicks, CTR, and cost per lead. By month 18, the account was generating north of 400 leads per month at a CPL that had dropped from around $180 to under $100.
Where The Numbers Fell Apart
The demand gen lead started comparing the Google Ads lead volume against the pipeline reports from Salesforce. Of those 400-plus monthly leads, fewer than 20 were becoming qualified opportunities. The close rate on Google Ads-sourced leads was under 2%. Meanwhile, leads from organic search and referral channels were closing at five to six times that rate.
The CEO asked a reasonable question: why are we spending $40K a month on a channel that generates volume but almost no revenue? The agency’s response was that lead quality was a sales problem, not a media problem. That answer is common. It is also wrong in this case.
The Diagnosis: Google Was Optimizing For The Wrong Outcome
The root cause was not keyword selection, ad copy, or landing page design. It was the conversion signal itself. The entire account was optimized around a single conversion action: lead form completions. Every campaign, every ad group, every bidding strategy was oriented toward getting Google to find more people willing to fill out a form.
Why Form Fills Are A Broken Signal For B2B
Google’s Smart Bidding algorithms are exceptionally good at finding patterns in data and then finding more users who match those patterns. When you tell the system that a form fill is a conversion, it learns what a form-filler looks like and goes hunting for more of them. The problem is that the characteristics of someone who fills out a form and the characteristics of someone who eventually buys enterprise software are often completely different populations.
In this account, a significant share of form fills were coming from students, consultants doing competitive research, and small businesses that would never qualify for the product. Google was doing exactly what it was told. It was doing it well. It was just told the wrong thing.
The PMax Problem
The account was also running Performance Max campaigns with broad audience signals and no job title or company size exclusions. PMax was generating high volume at low CPL, which made the agency’s reports look good, but nearly zero of those leads were converting downstream. This is a pattern covered in depth in Performance Max Vs Search Campaigns: Why Search-Only Wins For B2B Lead Gen, and it played out here exactly as expected.
No Offline Data In The Loop
The most critical gap: no offline conversion data was flowing back into Google Ads. The account had no visibility into what happened after a form was submitted. Google’s bidding algorithm could not distinguish between a lead that became a $40K deal and a lead that never responded to a single sales email. Without that feedback loop, Smart Bidding was flying blind.
This exact failure mode, where signal quality rather than campaign tactics is the root cause, is something we have documented in detail in How A B2B SaaS Team Fixed Google Ads Signal Quality And Recovered Pipeline.
The Fix: Rebuilding The Signal Stack Before Touching Campaigns
The instinct in most accounts is to start changing keywords, rewriting ads, or adjusting bids. Here, the first move was to fix the data layer. No campaign restructure matters if the algorithm is still optimizing for the wrong outcome.
Step 1: Salesforce Offline Conversion Sync
The team implemented Google’s offline conversion import via Salesforce, mapping the GCLID (Google click ID) captured at the point of form submission through to every subsequent stage in the CRM: MQL, SQL, opportunity created, and closed-won. This gave Google’s bidding system visibility into what happened weeks or months after the initial click.
For a detailed walkthrough of how this process works and why it changes everything, see How One SaaS Company Doubled Pipeline With Google Ads Offline Conversion Tracking.
Step 2: Redefining The Primary Conversion Action
Form fills were demoted from a primary conversion action to a secondary (observation-only) conversion. The new primary conversion action became “opportunity created” in Salesforce. This meant Google’s Smart Bidding would now optimize for finding users who ultimately become pipeline, not users who fill out forms.
Step 3: Importing Closed-Won Value
Closed-won deal values were imported as a value signal, enabling the eventual shift to target ROAS bidding on pipeline value. This step took longer to bear fruit because of the 60 to 90 day sales cycle, but it set the foundation for a bidding strategy that would get smarter over time.
Step 4: Scrapping PMax Temporarily
Performance Max was paused entirely. The reasoning was straightforward: PMax requires strong conversion signals to perform well in B2B. Feeding it form fills in a long-cycle B2B environment was generating noise, not signal. The plan was to revisit PMax once the offline conversion data had enough volume to give the algorithm something meaningful to work with.
Campaign Restructure: Collapsing 14 Campaigns Into 4
With the signal stack rebuilt, the team turned to campaign structure. The existing account had 14 campaigns, many of them overlapping, competing against each other in auction, and diluting the data that any single campaign could learn from.
The New Structure
The 14 campaigns were collapsed into four:
-
Non-brand search targeting bottom-of-funnel buyer intent keywords, scoped to job functions and company sizes that matched the ICP. Keywords like “workforce management software for midsize companies” replaced informational terms like “what is workforce management.”
-
Competitor search separating competitor brand terms into their own campaign with distinct messaging and bidding targets.
-
Brand search protecting branded terms with a dedicated campaign rather than letting them bleed into broad match elsewhere.
-
Remarketing limited to CRM-derived audiences of users who had visited high-intent pages (pricing, demo request) and matched ICP criteria from the CRM.
Audiences Built From CRM Data
Rather than relying on Google’s in-market audiences or affinity segments, the team built customer match audiences from Salesforce data: closed-won customers, qualified opportunities, and high-intent MQLs. These were used both as targeting layers and as similar audience seeds, giving Google a much tighter profile of who to pursue.
For broader guidance on how to structure SaaS accounts for pipeline outcomes, How To Structure Google Ads For SaaS Pipeline Growth In 2026 covers the framework in detail.
Bidding Strategy Shift: From Maximize Conversions To tROAS On Pipeline Value
The bidding strategy changed in two phases.
Phase 1: Conservative tCPA On Opportunities
For the first 30 days, the team ran target CPA bidding against the “opportunity created” conversion, setting the target conservatively to allow the algorithm to learn the new signal without throttling volume entirely. Lead volume dropped immediately, as expected. CPL nearly doubled. But the leads that did come through were qualitatively different: right job titles, right company sizes, right buying signals.
Phase 2: tROAS On Pipeline Value
After 45 days with enough opportunity-level data flowing, the team shifted to target ROAS bidding against closed-won deal value. This is where the compounding effect kicked in. Google could now see not just which clicks became opportunities but which clicks became revenue, and it could optimize for finding more clicks that looked like the high-value ones.
Understanding when tCPA versus tROAS is the right call is critical here. Google Ads Target CPA Vs Target ROAS: When Each Smart Bidding Strategy Works And When It Fails breaks down the decision framework. The short version: once you have reliable value data flowing back, tROAS almost always outperforms tCPA for B2B accounts with variable deal sizes.
The learning phase after the signal switch required patience. For teams navigating that transition, How To Exit Google Ads Learning Phase Faster And Protect Your Budget covers practical tactics to shorten the window without disrupting the algorithm.
The Results At 90 Days
By the end of 90 days, the account looked fundamentally different.
Pipeline contribution from Google Ads increased by roughly 80% on a flat budget. The total number of leads dropped significantly, from over 400 per month to around 120. But qualified opportunities went from fewer than 20 per month to over 35. Cost per lead went up. Cost per opportunity dropped sharply. The sales team reported that conversations with Google Ads leads were materially better: prospects were further along in their buying process, had clearer use cases, and were more likely to book follow-up calls.
The CEO stopped asking why Google Ads existed. The channel went from a questionable expense to the primary growth lever.
How groas Changes This Equation From Day One
This account spent 18 months and a significant amount of budget optimizing for the wrong signal before anyone caught it. That delay is not unusual. It is the default outcome when a traditional agency measures success by CPL and has no visibility into downstream revenue.
groas operates differently. The proprietary engine trained on over $500 billion in profitable ad spend identifies signal quality problems during the initial account analysis, not after 18 months of wasted spend. In a DWY (Done With You) engagement, a senior strategist works alongside your in-house team to build the offline conversion infrastructure, restructure campaigns around pipeline signals, and shift bidding strategies at the right pace while your team stays in control of execution. In a DFY (Done For You) engagement, groas owns the entire process end to end, from the CRM integration to the campaign rebuild to the bidding migration, including landing pages and offer structure.
The difference is not just speed. It is structural. A traditional agency assigns a media buyer who is managing 15 other accounts and whose performance is measured on surface-level metrics. groas pairs an engine that runs execution around the clock with a senior strategist who is accountable to pipeline and revenue, not lead volume. The signal rebuild that took this team weeks of internal work and agency negotiation is standard practice on every groas account from the start.
Month-to-month engagement, $0 onboarding, and no long-term contracts mean groas earns the next month by performing. If the pipeline numbers do not move, you walk.
What This Means For Your Account
The pattern in this case study is not rare. It is the default state of most B2B Google Ads accounts: optimizing for form fills, generating volume that does not close, and treating lead quality as someone else’s problem.
If you recognize any of these symptoms in your own account, here is the diagnostic checklist: Are you importing offline conversion data from your CRM into Google Ads? Is your primary conversion action a form fill or a downstream pipeline event? Is Performance Max running on broad signals without ICP-level filtering? Is your agency reporting on CPL without any visibility into close rates?
If the answer to any of those is yes, your account likely has a signal quality problem, and no amount of keyword optimization or ad copy testing will fix it.
For teams with in-house Google Ads knowledge who want to stay in the driver’s seat, groas DWY puts the engine and a senior strategist alongside your team. Get started today. For teams that want Google Ads fully handled from signal architecture to closed revenue, apply for groas DFY and the team will identify the right plan on the first call.
