A workflow-automation SaaS targeting mid-market operations teams in the US came to us spending $18,400 per month on Google Ads and generating an average of 4 sales-qualified leads per month. Their cost per SQL sat at $4,600, their sales team was frustrated, and the marketing director had been told the channel simply did not work for their ICP. Ninety days later, cost per SQL was $1,790, MQL-to-SQL conversion had improved from 9% to 24%, and the monthly SQL volume had risen from 4 to 11. Here is exactly what changed and why.
What the Audit Found
The account had been running broad-match keywords like "workflow software" and "process automation" without a meaningful negative keyword list. Google was routing spend toward SMB queries, IT helpdesk searches, and informational terms that had no commercial intent. Roughly 58% of clicks were coming from searches that would never convert into enterprise pipeline, which is a pattern we see repeatedly, and one we have written about in detail in our piece on why Google Ads stops generating quality leads.
The campaign structure was also flat: one campaign, four ad groups, no segmentation by job title intent, product use case, or funnel stage. Every click landed on the same homepage. There was no dedicated landing page for any specific ICP segment, and the page itself had no social proof targeted at operations leaders, only generic product screenshots aimed at developers. Attribution was last-click inside Google Ads, so the team had no visibility into which keywords were actually contributing to closed revenue.
Phase 1: Restructure and Negative Keyword Cleanup (Days 1-30)
We rebuilt the campaign structure around three clear intent clusters: direct competitor comparisons, job-function-specific pain points ("operations manager workflow bottlenecks"), and integration-intent searches ("Zapier alternative for enterprise"). Each cluster got its own campaign with separate budgets, match types, and bid strategies. We added 340 negative keywords on day one, covering SMB modifiers, job titles outside the ICP, and informational prefixes like "what is" and "how does." For a detailed walkthrough of the negative keyword methodology, see our guide on eliminating wasted spend with negative keywords.
Within the first 30 days, impression share on irrelevant queries dropped by 71% and average CPC fell from $14.20 to $9.80, even though we had not changed bids directly. Removing low-quality traffic improved Quality Scores across the board, which in turn lowered CPCs automatically. Total click volume dropped 34%, but the clicks that remained were far more qualified.
Phase 2: Landing Page Segmentation (Days 15-45)
We built three dedicated landing pages, one per intent cluster, each with a headline that matched the search query theme, a single conversion action (demo request), and social proof specific to operations leaders: a quote from a VP of Operations at a 400-person logistics firm, an ROI stat ("reduced manual process time by 11 hours per week"), and a single case study summary. The homepage had none of this. It was trying to serve developers, executives, and SMBs simultaneously, which meant it served none of them well. HubSpot's research on landing page conversion rates consistently shows that message-matched pages outperform generic destinations by a significant margin, and our results confirmed that pattern.
Landing page conversion rate across the three new pages averaged 6.8%, compared to 1.4% on the homepage. That single change - sending qualified clicks to a relevant, specific page instead of a generic homepage - accounted for roughly half of the total SQL volume improvement. If your paid search clicks are not converting, the problem is often not the ads themselves. We have covered this in more depth in our article on why your B2B landing page does not convert.
Phase 3: Attribution Fix and Budget Reallocation (Days 30-90)
We moved the client off last-click attribution and onto a data-driven model inside Google Ads, then connected offline conversion imports from their CRM so that closed-won deals and SQL-stage pipeline events were flowing back into the bidding algorithm. This is a step most B2B advertisers skip, and it is costly: without it, Smart Bidding optimises toward form fills, many of which are low-quality MQLs or even spam. With CRM data feeding back in, the algorithm learned which keyword patterns were producing pipeline, not just contact form submissions.
By day 60, we had 45 days of data showing clear cost-per-SQL differences between the three campaign clusters. The competitor-comparison cluster had a cost per SQL of $1,200, the job-function cluster was at $1,900, and the integration-intent cluster was at $3,400. We reallocated 30% of the integration-intent budget into the competitor cluster and reduced daily caps on the weakest ad groups. Google's Smart Bidding documentation explains how offline conversion imports interact with Target CPA strategies, and we followed that setup precisely.
The Final 90-Day Numbers
- Cost per SQL: $4,600 at start, $1,790 at day 90 (61% reduction)
- Monthly SQL volume: 4 at start, 11 at day 90
- MQL-to-SQL conversion rate: 9% to 24%
- Average CPC: $14.20 to $9.80
- Landing page CVR: 1.4% (homepage) to 6.8% (segmented pages)
- Total monthly ad spend: held flat at $18,400
Total monthly spend did not increase by a single dollar. Every improvement came from structural changes: better keyword segmentation, a real negative keyword strategy, message-matched landing pages, and attribution data that actually reflected business outcomes. The channel had not stopped working. It had never been set up correctly in the first place.
What to Take From This
The three levers that moved the numbers most were negative keywords, landing page specificity, and offline conversion imports. Of these, the landing page change produced the fastest impact - visible within the first two weeks. The attribution fix produced the most durable impact, because it compounded over time as the bidding algorithm accumulated better data. Structural campaign rebuilds like this typically take 60-90 days to fully stabilise, so the temptation to pause or cut budget at the 30-day mark - before the algorithm has learned - is one of the most common and damaging mistakes B2B teams make.
If your paid search account has been running for more than six months without a structural audit, the numbers above suggest the opportunity cost is significant. A flat $18,400 monthly budget going from 4 SQLs to 11 SQLs per month, at the same spend, is the equivalent of adding $55,200 in monthly pipeline generation capacity without increasing the budget at all.