I want to walk through a real engagement from start to finish, because the details matter more than the headline number. A 55% CPA reduction sounds impressive in a pitch deck, but the interesting part is how it happened — what we found, what we changed, and what we learned along the way.
Some specifics have been anonymized at the client’s request, but the numbers, timeline, and methodology are accurate.
The Starting Point
The client is an enterprise SaaS company selling a workflow automation platform. Their target buyers are operations leaders at mid-market and enterprise companies — think VP of Operations at a 500-2000 person company.
When they came to us, they’d been running Google Ads and LinkedIn Ads in-house for about 18 months. Their monthly ad spend was around $45,000 across both platforms. They were generating leads, but the cost per qualified lead had been climbing steadily for six months and was sitting at $380.
Their sales team reported that lead quality had also been declining. Close rates had dropped from 18% to about 11% over the same period. So they were paying more per lead AND each lead was less likely to close. Not a great combination.
Their goal was simple: get CPA back under $200 while maintaining or improving lead quality. They gave us 90 days to prove the approach.
Week 1-2: The Audit
Before we changed a single setting, we spent two weeks in their accounts, their CRM, and their analytics. Here’s what we found.
Google Ads problems:
The account had 14 campaigns with overlapping keywords. Several campaigns were bidding against each other for the same search terms. This is a surprisingly common issue in accounts that have grown organically over time — someone creates a new campaign for each product feature or use case, and nobody cleans up the overlap.
The search terms report was ugly. About 28% of Google spend was going to search terms that were either irrelevant (informational queries, competitor employee searches) or too broad to indicate purchase intent. We found clicks from terms like “workflow automation free template,” “what is workflow automation,” and “[competitor name] careers.”
Their landing pages were generic. All Google Ads traffic pointed to either the homepage or a single “Request a Demo” page. The demo page had a 2.1% conversion rate, which sounds okay until you realize they were paying $14-18 per click for enterprise SaaS keywords. At $16 average CPC and 2.1% conversion rate, the math produces a $762 cost per demo request — before even considering that many demo requests never show up.
LinkedIn Ads problems:
They were running sponsored content campaigns targeting a broad audience: anyone with “VP” or “Director” in their title at companies with 200+ employees. The targeting was directionally right but way too broad. The content was product-focused — essentially feature announcements — which doesn’t work well on LinkedIn for cold audiences.
Their LinkedIn CPA was $520. Painful at any scale, but they were spending $15K/month on it.
CRM and analytics problems:
Conversion tracking was partially broken. The Google Ads pixel was counting both “demo request submitted” and “contact us submitted” as the same conversion event, so they couldn’t distinguish high-intent from low-intent leads in the ad platform. Their CRM had lead source attribution, but it wasn’t flowing back to the ad platforms for optimization.
Week 3-4: The Restructure
We didn’t start optimizing. We rebuilt.
Google Ads restructure:
We consolidated the 14 campaigns into 5, organized by intent level:
- Brand search — people searching for the company name (previously spread across 3 campaigns)
- High-intent non-brand — searches indicating active buying intent (“workflow automation software,” “best workflow automation for enterprises,” “[competitor] alternative”)
- Mid-intent non-brand — searches with commercial intent but earlier in the funnel (“workflow automation tools comparison,” “how to automate business workflows”)
- Competitor campaigns — targeting competitor brand names
- Remarketing (RLSA) — previous site visitors searching related terms
We built a negative keyword list of 340+ terms and applied it across all campaigns. We switched from broad match to a mix of exact and phrase match on high-intent campaigns.
Landing pages:
We built three dedicated landing pages:
- One for high-intent searches, focused on pricing, demo scheduling, and competitor comparison
- One for mid-intent searches, offering a free workflow assessment rather than pushing straight to a demo
- One for competitor campaigns, with a direct comparison angle
Each page had a single CTA, a fast load time (under 2 seconds), and messaging that matched the search intent of its target campaign.
LinkedIn restructure:
We narrowed targeting significantly. Instead of all VPs/Directors at 200+ employee companies, we focused on:
- Specific job titles: VP Operations, Head of Operations, Director of Business Process, Head of Digital Transformation
- Company size: 500-5000 employees (their sweet spot based on CRM data)
- Industries where they had the best close rates: financial services, healthcare, and manufacturing
We replaced the product-focused content with a lead magnet approach: a “Workflow Automation Maturity Assessment” that targeted the pain points their best customers had described in sales calls.
Tracking fixes:
We separated conversion events: demo request became the primary conversion, contact form became a secondary event with lower value. We set up offline conversion import from their CRM so that Google and LinkedIn would eventually learn which leads actually became sales opportunities, not just which leads filled out a form.
Month 2: Optimization
With the new structure in place and clean data flowing, we moved into optimization.
Google Ads adjustments:
We switched the high-intent campaigns to Target CPA bidding with an initial target of $250 (above their goal, but we wanted to give the algorithm room to learn). As data accumulated, we tightened the target in $20 increments every two weeks.
We ran A/B tests on ad copy. The biggest finding: ads that included a specific number (“Used by 200+ enterprise teams” or “30-day implementation”) consistently outperformed generic value proposition ads. Specificity wins.
We added dayparting, reducing bids by 60% between 8 PM and 6 AM and on weekends. For B2B enterprise, these were dead hours — clicks happened but conversions didn’t.
LinkedIn adjustments:
The lead magnet approach worked immediately. CPA dropped from $520 to $310 in the first month with the new creative. We then tested different angles for the assessment offer and found that emphasizing “find out how your workflow automation compares to industry benchmarks” outperformed “get a free assessment” by about 40% on conversion rate.
We also implemented LinkedIn’s website retargeting to show ads specifically to people who had visited the pricing page or case studies on the website but hadn’t converted.
Landing page optimization:
We ran 4 A/B tests on the high-intent landing page over 6 weeks:
- Headline test: “Automate Your Workflows in 30 Days” beat “Enterprise Workflow Automation Platform” by 34% on conversion rate
- Form length test: 4 fields beat 7 fields by 28% on submissions (we added qualification questions later in the funnel)
- Social proof test: Specific client logos with results (“Company X reduced manual processes by 60%”) beat generic “trusted by 500+ companies” by 22%
- CTA test: “See How It Works” beat “Request a Demo” by 18% — the softer CTA reduced friction
Month 3: Results
By the end of the 90-day period, here’s where things stood:
CPA: $171 (down from $380 — a 55% reduction)
That number breaks down to:
- Google Ads CPA: $148 (down from $312)
- LinkedIn Ads CPA: $224 (down from $520)
- Blended CPA across both: $171
Lead quality: Sales-qualified lead (SQL) rate improved from 11% to 19%. This is arguably more important than the CPA reduction. The combination of better targeting, intent-based landing pages, and offline conversion feedback meant the algorithms were optimizing toward leads that actually progressed through the sales pipeline.
Revenue impact: With a lower CPA and higher SQL rate, the effective cost per SQL dropped from $3,455 to $900. At their average deal size, this shifted paid marketing from a questionable investment to a clearly profitable channel.
What the monthly ad spend looked like:
Total spend stayed roughly the same at $42K/month. We reallocated budget from low-performing campaigns and geographies to the high-performers. The same spend was simply producing 2.2x more qualified leads.
What Actually Made the Difference
If I had to rank the changes by impact, here’s how I’d stack them:
- Landing pages matched to intent — this was probably 30-35% of the improvement by itself. Sending the right person to the right page with the right message is the single highest-leverage thing in paid media.
- Cleaning up wasted spend — eliminating irrelevant search terms and narrowing LinkedIn targeting recovered roughly 25% of the budget, which we reinvested in proven campaigns.
- Conversion tracking fixes + offline import — getting the algorithms optimizing toward actual sales opportunities rather than form fills improved lead quality dramatically over time.
- Landing page A/B testing — the cumulative effect of four winning tests improved the landing page conversion rate from 2.1% to 4.8%.
- Campaign restructure and bid strategy — the cleaner structure allowed smarter budget allocation and better algorithmic optimization.
None of these are revolutionary tactics. There’s no secret here. It’s systematic work: audit, restructure, fix the data, optimize against the right goals, and test continuously. The difference between a $380 CPA and a $171 CPA wasn’t one brilliant insight. It was twenty good decisions executed consistently.
Twelve Months Later
The client is still with us. Their CPA has continued to improve — it’s now hovering around $145. More importantly, the sales team reports that lead quality has sustained at the improved level, and close rates are now at 21%, above their historical best.
We’ve expanded into YouTube prospecting and are testing programmatic retargeting, using the profitable Google and LinkedIn campaigns as the core engine.
The total pipeline value attributed to paid media has increased by 280% from when we started, on roughly the same monthly ad spend. That’s the power of efficiency: you don’t always need more budget. You often just need to stop wasting the budget you have.
Think your ad accounts have room for improvement? Request a free audit and we’ll tell you exactly what we’d change.