TLDR;
- Our Strategic Approach: We built a multi-stage advertising funnel (TOFU/MOFU/BOFU) on Meta, using a rigorous testing framework to systematically identify winning locations, audiences, and ad creatives.
- Verified Results: Generated 4,622 B2B software registrations at an average Cost Per Registration of just $2.38.
- Core Takeaway: For early-stage SaaS acquisition, broad "Worldwide" targeting can dramatically lower costs, and Lookalike Audiences built from high-intent user actions (like trials and purchases) still outperform most interest-based targeting.
Driving signups for a B2B software product is often a pricey game. You’re usually told to expect high costs per acquisition, especially when you’re just starting out. But what if you could get thousands of new users registering for your software for less than the price of a coffee? In this breakdown, I'm going to walk you through exactly how we did that for a client, generating 4,622 software registrations at an average cost of just $2.38 each.
The client was Three Calls, and their product is Feedback W, a software tool designed to help Amazon sellers improve their feedback scores and streamline review requests. It’s a classic B2B SaaS play: a specific tool for a specific professional audience. The goal was simple – get more Amazon sellers to register for the software, but to do it efficiently without burning through cash.
A lot of agencies would jump straight into Ads Manager, throw a few audiences together, and hope for the best. That’s a recipe for disaster. Great campaigns aren't about luck; they're about engineering a system. Our approach always starts with a detailed strategy plan, long before we spend a single dollar on ads. It's about mapping the entire customer journey, from the first ad they see to the moment they become a paying customer. This case study isn't just about the final numbers; it’s about the methodical process that got us there.
The Strategic Foundation: Planning Before Spending
Most advertising campaigns fail before they even launch. They fail because there's no coherent strategy. People get seduced by the idea of quick results, pick a few targeting options that *feel* right, and set their campaigns live. The result is almost always a slow, expensive bleed of the marketing budget with very little to show for it.
We operate differently. Every project, including this one for Feedback W, begins with a thorough analysis of strategy. We map out the entire advertising funnel, research the competitive landscape, identify the most promising audiences, and define a clear testing roadmap. This isn't just busywork; it's the blueprint that ensures every action we take is deliberate and measurable. For this client, the plan was the difference between spending $11,000 to get a handful of overpriced leads versus spending it to acquire over 4,600 new users.
Our initial phase involved a thorough review of what other players in the Amazon seller tool space were doing. What kind of ads were they running? What angles were they using? What offers were they pushing? This gave us a baseline understanding of the messaging that was already resonating with our target audience—Amazon sellers. But more importantly, it showed us where the gaps were, giving us opportunities to say something different and cut through the noise.
With that context, we designed a multi-stage funnel specifically for acquiring SaaS users. It’s a structure we use often because it works. It aligns with a potential customer’s natural journey from being unaware of a solution to becoming an active, paying user. Trying to push a cold prospect straight to a sale is like asking for marriage on a first date; you have to build the relationship first.
SaaS User Acquisition Funnel Architecture
The funnel we mapped out had three clear stages:
- Top of Funnel (TOFU): This is all about prospecting. The goal here is to reach a broad but relevant audience of Amazon sellers who have likely never heard of Feedback W. The main objective is to get them to click through to the website and register for an account. This is where the bulk of the ad spend goes.
- Middle of Funnel (MOFU): This stage is for retargeting. We’re going after people who visited the website from our TOFU ads but didn't sign up. Maybe they got distracted, or maybe they weren't convinced yet. MOFU ads serve as a reminder, reinforcing the value proposition and nudging them back to complete their registration.
- Bottom of Funnel (BOFU): This is the final stage, focused on activation and conversion. We target users who have already registered for a trial but haven’t become active or upgraded to a paid plan. These ads aim to keep them engaged during their trial period, highlight key features, and ultimately encourage them to make a purchase.
With the funnel structure defined, the next part of the plan was to outline our targeting and testing strategy. For a new account, we always start by testing a wide range of detailed targeting options—interests, behaviours, and demographics that align with the ideal customer. For Feedback W, our research pointed to audiences like 'small business owners', 'retail page admins', and people with interests in 'Shopify', 'Amazon Marketplace', and 'E-commerce'.
However, the plan was always to move beyond basic interests as quickly as possible. The real power on platforms like Meta comes from Lookalike Audiences. So, our strategy detailed a clear progression: use initial interest-based campaigns to drive traffic and gather data, then use that data to build high-quality Lookalike Audiences. We planned to create Lookalikes based on website visitors, completed registrations, trial users, and eventually, paying customers. The idea is simple: the further down the funnel someone is, the more valuable they are, and the better a Lookalike Audience built from them will be.
Finally, the strategy plan laid out a rigorous optimization framework. We weren’t just going to 'see what works'. We detailed specific tests we wanted to run, including location targeting (Worldwide vs. US vs. Developed Countries), audience performance (Lookalikes vs. Interests vs. Broad), and of course, continuous split-testing of ad creatives and copy. Every element was designed to be a variable we could test, measure, and optimize. This upfront planning meant that once the campaigns launched, we had a clear set of instructions to follow, allowing us to make data-driven decisions quickly and efficiently.
Executing the Funnel: Campaign & Targeting Overview
With the strategy in place, it was time for execution. We structured the Meta Ads account to mirror our funnel, creating separate campaigns for TOFU, MOFU, and BOFU. This separation is necessary; it allows you to allocate budget effectively and tailor your messaging to the user's level of awareness. You don't talk to a complete stranger the same way you talk to someone who has already started a trial of your software.
One of the first and most impactful tests we ran was on location targeting. It’s common for B2B SaaS companies, especially those based in the US or UK, to focus their ad spend exclusively on so-called 'Tier 1' developed countries. The logic is that these markets have more purchasing power. While often true, it’s also a lazy assumption that can cost you a fortune. We decided to test this head-on.
We set up three distinct top-of-funnel prospecting campaigns, all with the same audiences and ads, but each targeting a different geographic cluster: one for the US only, one for a list of 'Developed Countries', and one targeting 'Worldwide'. The results were, to put it mildly, dramatic.
Location Targeting Test Results
A head-to-head comparison of Cost Per Registration (CPR) across different geographic targeting strategies revealed a clear winner.
| Campaign (Location Targeting) | Registrations | Amount Spent | Cost Per Registration (CPR) | Decision |
|---|---|---|---|---|
| TOFU (US Only) - Complete Registration | 24 | $925.60 | $38.57 | Turn Off |
| TOFU (Developed Countries) - Complete Registration | 232 | $1,337.42 | $5.76 | Monitor / Reduce |
| TOFU (Worldwide) - Complete Registration | 1,083 | $336.75 | $0.31 | Scale |
The US-only campaign was an absolute cash fire, bringing in registrations at a staggering $38.57 each. The 'Developed Countries' campaign was better but still not great, at $5.76 per registration. But the 'Worldwide' campaign was a revelation, delivering over 1,000 registrations at an incredible $0.31 each. It was a no-brainer. We immediately turned off the US campaign and started funnelling the budget towards the worldwide strategy.
This is a classic example of challenging assumptions. While the trial-to-paid conversion rate from lower-income countries might be lower, the sheer cost-efficiency of acquiring the initial registration at the top of the funnel more than made up for it. We could acquire dozens of users from around the world for the price of one from the US. This allowed us to feed our funnel with a massive volume of users, gather data rapidly, and let our MOFU and BOFU campaigns work on converting the most engaged ones.
Next up was audience testing. As per our plan, we began with detailed interest targeting but quickly transitioned to testing Lookalike Audiences as soon as we had enough conversion data. Inside our prospecting campaigns, we ran multiple ad sets, each targeting a different audience, to see which would perform best. The results provided another clear lesson. While some interest-based audiences, like 'small business owners', performed reasonably well, they were consistently outperformed by our Lookalike Audiences.
We tested everything: broad targeting (which performed poorly), detailed interests around e-commerce and online marketplaces (also poor performers), and a suite of Lookalikes. The data showed that Lookalikes built from high-intent actions were the clear winners. Specifically, a 1% Lookalike of users who had previously completed registration and a 1% Lookalike of previous purchasers delivered the best performance. This confirms a long-standing principle of paid advertising: the algorithm is far better at finding people similar to your best customers than you are at guessing their interests.
Audience Performance Hierarchy
Tier 1: High-Performance Lookalikes (Winners)
Lookalikes of high-intent actions consistently delivered the lowest Cost Per Registration. This included 1% Lookalikes of Completed Registrations, Website Visitors, Trials, and Purchases. These became our core scaling audiences.
Tier 2: Specific Detailed Targeting (Viable)
A small number of highly relevant detailed targeting options, such as 'Small Business Owners', proved to be effective and provided a solid, if smaller, source of registrations.
Tier 3: Broad & General Interests (Turned Off)
Broad targeting and general interest categories like 'Online Retail' or 'Marketplace Interests' performed poorly. They were too generic, leading to high spend with few conversions, and were quickly disabled.
Our optimization process was relentless. We monitored performance daily. Any audience that wasn't hitting our target CPR after a reasonable amount of spend was shut off. This ruthless approach freed up budget to be reinvested into the winners—the Worldwide location targeting and the high-intent Lookalike audiences. This is what active campaign management looks like. It’s not 'set and forget'; it's a continuous process of testing, learning, and reallocating resources to what's proven to work.
Creative Testing: Finding the Message That Converts
Getting your targeting right is half the battle, but it's worthless if your ads are rubbish. Your ad creative and copy are what do the actual work of persuasion. For Feedback W, we knew we had to grab the attention of busy Amazon sellers scrolling through their feeds and quickly communicate why this tool was worth their time. We applied the same methodical testing process to our creatives as we did to our targeting.
We developed a variety of ad angles, each with different taglines and body copy. We didn't just guess what would work; we created hypotheses based on potential pain points for Amazon sellers. Some ads focused on the benefit of "boosting your Amazon sales," others on the efficiency of "automating your Amazon feedback," and others on the ultimate goal of "unlocking Amazon success." Each ad was a small experiment designed to see which message resonated most.
We ran these different ad variations across all our winning ad sets. A common mistake people make is to judge an ad's performance based on how it does in just one ad set. That's a flawed approach because performance can vary. Instead, we look at the aggregate performance of an ad across the entire campaign. This gives us a much more reliable picture of which creatives are truly the strongest.
The process was simple but effective: launch a batch of ads, let them run for a few days to gather enough data, and then analyse the results. We looked at metrics like click-through rate (CTR), cost per click (CPC), and, most importantly, the number of registrations and the cost per registration. Any ads that were clear underperformers—those with low engagement and high costs—were switched off. The budget was then automatically reallocated by the platform to the remaining, better-performing ads. This continuous cycle of testing and culling ensures that your ad spend is always being concentrated on the creatives that are most effective at driving results.
Final Campaign Performance Snapshot
The final results after months of systematic optimisation, highlighting the overall efficiency and the impact of our rigorous testing process.
One creative, with the headline "Unlock amazon success," emerged as a clear winner, driving over 633 registrations on its own. This iterative process of refinement is how you go from mediocre results to outstanding ones. It's not about finding one 'magic' ad on day one. It's about building a system that allows you to consistently test, learn, and improve over time, ensuring your campaign's performance gets better and better.
It's worth noting that later in the project, the client had to reduce their budget, which is why you see only a few campaigns active in the final account snapshot. However, the system we built was designed for this. Because we had already identified the top-performing campaigns, audiences, and creatives, we could simply pause the lower-priority tests and concentrate the reduced budget on the proven winners, maintaining efficiency even at a lower spend. This is the hallmark of a robust advertising system—it's resilient and adaptable to changing business needs.
The Final Tally: A System That Converts
After implementing our strategy and running through cycles of optimisation, the results speak for themselves. Across all campaigns, we spent a total of $11,013.42 and generated 4,622 software registrations. This works out to an average cost per registration of just $2.38—a phenomenal result for a B2B SaaS product.
The success of this project wasn't down to a single secret trick or a lucky break. It was the direct result of a systematic, data-driven approach. We didn't just run ads; we engineered an entire acquisition funnel.
Let’s review the core pillars that made this work:
- Strategy First: We started with a comprehensive plan that defined the funnel, identified target audiences, and laid out a clear testing roadmap. This preparation was the foundation for everything that followed.
- Rigorous Testing: We challenged common assumptions by testing everything. The location targeting test alone, which proved 'Worldwide' was dramatically more efficient than 'US Only', was a massive win that saved the client thousands of dollars and unlocked massive scale.
- Data-Driven Audience Selection: We let the data guide our targeting. While we started with interest-based audiences, we quickly pivoted to what was working: Lookalike Audiences built from high-intent user data. We were ruthless in cutting underperforming audiences to focus the budget on what was proven to work.
- Iterative Creative Optimisation: We continuously split-tested ad copy and messaging, analysing aggregate performance to identify the ads that truly resonated with Amazon sellers and turning off those that didn't.
This project is a perfect illustration of our philosophy: great campaigns fail without expertise. It's not enough to know your way around Ads Manager. You need a strategic framework to guide your decisions, an understanding of how to build and optimise a full-funnel system, and the discipline to follow the data, even when it leads to uncomfortable or counter-intuitive conclusions.
For many SaaS businesses, particularly in the B2B space, achieving this kind of efficiency in user acquisition can feel out of reach. They get stuck with high costs per lead, struggle to scale their campaigns profitably, and waste budget on strategies that were never going to work. Often, the missing piece is the expertise to design and execute a system like the one we built for Feedback W.
If you're running a business and finding that your advertising efforts aren't delivering the tangible revenue growth you need, it might be time to bring in a specialist. If you're interested in a strategy like this for your own business, we can help. We offer a free strategy consultation where we can take a look at your ad account together and figure out the right approach to start driving real results.
Cite this Case Study
If referencing this research, data, or strategy in your publication, please use the attribution below.
Lukas Holschuh
Founder, Growth & Advertising Consultant
Great campaigns fail without expertise. Lukas and his team provide the missing strategy, optimizing your entire advertising funnel—from ad creatives and copy to landing page design.
Backed by a proven track record across SaaS, eLearning, and eCommerce, they don't just run ads; they engineer systems that convert. A data-driven partnership focused on tangible revenue growth.