MetricsMatter: 2.5x ROAS for SaaS in 2026

Here’s the MetricsMatter 5.0 campaign teardown, where we show how measurement turns marketing spend into actual results and proves communications ROI is a real, measurable number. We just finished a deep dive on a B2B SaaS product launch that makes the case perfectly: a three-month paid ad campaign to get sign-ups for an analytics platform. The goal was pure direct response, getting qualified leads to become trial users. We had to track every dollar and be ready to pivot based on what the data told us.

Campaign Overview: Precision Targeting for a Niche SaaS

We ran the “InsightEngine Launch” campaign from July 1 to September 30, 2026. The only goal was to get new trial users for an AI analytics platform built for mid-market finance companies. We put a $75,000 budget into Google Ads (Search and Display) and carved out a smaller piece for LinkedIn Ads to hit specific job titles. The conversion event was a completed sign-up for the 14-day free trial, which asked for detailed company info, so we knew the leads were serious. Our whole strategy was hyper-segmentation. We were laser-focused on finding Chief Financial Officers, Heads of Analytics, and Senior Data Scientists inside companies with $50 million to $500 million in revenue, because those were the exact people who would actually use the platform’s predictive modeling features. For the creative, we used problem-solution messaging. We skipped the generic “boost your insights” stuff and wrote ads that hit on real pain points like “Eliminate manual data reconciliation” or “Forecast market shifts with 95% accuracy.” For display ads, we had custom graphics made to show the actual platform UI, ditching stock photos entirely. This approach conveyed sophistication and showed exactly what the tool did.

Initial Performance: Baseline Metrics and Early Surprises

Month one was all about setting a baseline. We got a lot of impressions, sure, but the conversion rate was definitely lower than we wanted. Here’s how the numbers broke down: Month 1 Performance (July 2026) | Metric | Google Search | Google Display | LinkedIn Ads | Total |
| :, , | :, | :, – | :, – | :, |
| Budget Allocated | $25,000 | $10,000 | $5,000 | $40,000 |
| Impressions | 1,200,000 | 800,000 | 150,000 | 2,150,000 |
| Clicks | 48,000 | 4,000 | 1,200 | 53,200 |
| CTR | 4.00% | 0.50% | 0.80% | 2.47% |
| Conversions | 120 | 8 | 10 | 138 |
| Conversion Rate | 0.25% | 0.20% | 0.83% | 0.26% |
| Cost Per Click (CPC) | $0.52 | $2.50 | $4.17 | $0.75 |
| Cost Per Lead (CPL) | $208.33 | $1250.00 | $500.00 | $289.86 | That initial Cost Per Lead (CPL) of almost $290 was way over our $150 target. Google Search brought in most of the conversions, but its CPL was still too high. LinkedIn Ads had a pricier CPC but its conversion rate was much better, telling us the traffic quality was there. And the Google Display Network, as expected, had an awful CPL, which meant our targeting or creative was missing the mark completely on that channel. The big surprise from the first month’s data was the weird split between desktop and mobile. Mobile impressions made up 40% of our total, but the CTR was 20% lower than desktop. But then we saw that the mobile conversion rate was actually 15% *higher*. This suggested that the smaller group of people who clicked on mobile were far more serious about signing up, probably because they could do it right then and there. That finding told us exactly where to focus our optimization efforts next: mobile.

Optimization Phase: Iteration and Data-Driven Adjustments

With the July data in hand, we went on an optimization tear through all of August.

Targeting Refinements

On Google Search, we killed most of our broad match keywords and switched almost entirely to exact match and phrase match. We went after long-tail queries like “AI predictive analytics for financial planning” which cut down on garbage impressions and clicks, instantly improving our traffic quality. We also got ruthless with negative keywords, blocking anything with “free analytics tools” or “personal finance software” to keep out the wrong crowd. Over on LinkedIn, we layered job titles with company size and industry to narrow our audience, and we started using Matched Audiences more by uploading target account lists to get right to the decision-makers.

Creative Overhaul and A/B Testing

Our display ads were clearly not working, so they got a total overhaul. We ran a bunch of A/B tests:

• Headline Variation: A direct, benefit-driven headline like “Achieve 95% Forecast Accuracy” beat a vague one like “Unlock Your Data’s Potential” by an average 30% in CTR every single time.
• Call-to-Action (CTA): “Start Free Trial” gave us a 15% higher conversion rate than “Learn More” across the board which confirmed we were right to focus on direct response.
• Visuals: We tested short, 15-second animated videos of a platform feature against our static images on the Google Display network, and the videos got a 22% higher engagement rate (defined as clicks or 3-second view duration).

Creative optimization was a continuous process. We were in the accounts every week, rotating ad variations, killing the losers, and putting more budget behind the winners.

Bid Strategy Adjustments

We started with manual bidding just to get our bearings and collect some data. Once we had enough conversion history by month two, we switched Google Ads over to an automated Target CPA (Cost Per Acquisition) strategy, starting with a target of $180. This let the algorithm take over the bid adjustments to hit our CPA goal, and it freed us up to work on other things. For LinkedIn, we went with Maximum Delivery within a budget cap, since its algorithm often does better when you give it room to find conversions on its own. Based on that early mobile data, we made a big adjustment. Even with the lower CTR, the higher conversion rate was too good to ignore, so we jacked up our mobile bids by 20% on Google Search to capture more of that high-intent traffic. At the same time, we put resources into fixing the mobile landing page experience by simplifying the sign-up form and making sure the page loaded instantly. In 2026, a fast mobile experience is table stakes.

Results: Tangible ROI from Communications Data

By the end of the campaign, all these optimizations were clearly paying off. Here’s what the numbers looked like for September: Month 3 Performance (September 2026) | Metric | Google Search | Google Display | LinkedIn Ads | Total |
| :, , | :, | :, – | :, – | :, |
| Budget Allocated | $15,000 | $5,000 | $15,000 | $35,000 |
| Impressions | 800,000 | 300,000 | 250,000 | 1,350,000 |
| Clicks | 40,000 | 1,800 | 2,000 | 43,800 |
| CTR | 5.00% | 0.60% | 0.80% | 3.24% |
| Conversions | 180 | 12 | 40 | 232 |
| Conversion Rate | 0.45% | 0.67% | 2.00% | 0.53% |
| Cost Per Click (CPC) | $0.38 | $2.78 | $7.50 | $0.80 |
| Cost Per Lead (CPL) | $83.33 | $416.67 | $375.00 | $150.86 | We actually cut the budget in the final month to $35,000 because our spending got so much more efficient. We pulled way back on Google Display, its CPL just never got good enough, and moved that money over to LinkedIn where we were getting much better leads. So for the whole three-month campaign, we spent the full $75,000 and got 450 trial sign-ups. Our blended Cost Per Lead (CPL) went from a painful $289.86 in month one all the way down to $150.86 in month three, hitting our target right on the nose. The overall campaign conversion rate doubled from 0.26% to 0.53%. But the number that really mattered to the business was Return on Ad Spend (ROAS). Looking at our internal data, we knew about 15% of trial users would become paying customers within six months, and the average Customer Lifetime Value (CLTV) was $2,500. Here’s the math:
Calculation:
Total Conversions: 450
Paying Customers (15% of conversions): 450 * 0.15 = 67.5 (we’ll use 67 for the math)
Total Revenue: 67 * $2,500 = $167,500
Total Ad Spend: $75,000
ROAS: $167,500 / $75,000 = 2.23x
That 2.23x ROAS showed a solid return. For every dollar we put into ads, we got $2.23 back in revenue, proving the ROI was there. The InsightEngine Launch campaign showed that obsessive tracking, constant A/B testing, and moving the budget around based on performance can turn a marketing cost center into a revenue driver. It’s about spending smarter.

Lessons Learned and Future Implications

We took away a few key lessons from the InsightEngine campaign. First, initial performance numbers don’t tell you the full story. That high CPL in month one could have easily scared us into pausing channels that ended up working, but we stuck with the optimization plan. Second, data granularity is everything. Finding that mobile conversion rate discrepancy was the turning point. If we had just looked at the low CTR, we would have cut mobile bids and missed out on our most motivated users. Third, automated bidding really does work better than manual for hitting a specific CPA, but only after you’ve fed it enough conversion data to learn from. As a quick aside, just turning on “smart” bidding without good conversion data and a clear goal is a terrible idea. It’s like telling a self-driving car to just “go” without a destination, you have no idea where you’ll end up. For the next campaign, we’re planning to build in more predictive analytics from the start. We want to use machine learning models that can help forecast keyword and audience performance so we can make proactive adjustments. We’re also going to dig into more advanced attribution models instead of just relying on last-click. Using something like data-driven attribution should give us a much better sense of how each touchpoint contributes to a conversion. This will give us an even sharper view of our communications ROI and help us fine-tune the marketing mix. That constant feedback loop, analyzing the data, executing changes in the campaign, and then analyzing the new data, is what makes or breaks a digital marketing effort. It’s a dynamic, iterative cycle. You have to be in the data constantly and be willing to change your plan based on what the numbers are telling you.