5 Ad Tests Every SaaS Should Be Running (But Most Don’t)

Most SaaS teams run ads like a vending machine: put budget in, hope demos come out. But when results disappoint, the usual adjustments are superficial…change the colour, swap out the stock image, maybe tweak the call-to-action. But you can’t expect improvements when you’re making random adjustments; you need take a structured approach and split test.

Maybe you’ve overlooked this up until now, but Ads platforms aren’t just a way to buy traffic…They’re also the most efficient digital laboratory for understanding what your market responds to. Run the right tests and you’ll uncover insights about your messaging, positioning, and audience that make every pound you spend sweat.

1) Pre-Qualifying Ad Copy

Most SaaS CMO’s run ad campaigns designed to get as many clicks as possible, but you don’t want the most clicks, you want to protect your budget and use it to attract relevant clicks.

To help improve this, a strategy I love to deploy is “pre-qualifying” ad copy. This involves using ad copy filters before you get charged for the click. Done well, it saves budget and improves lead quality. You can use whole range of pre-qualifiers, such as ICP call-outs “Founder” or “IT Manager”, or by including a price tag on the Ad “$199 p/m”, or even try using long-form filters like “Not Sure How to Invest a £10k IT Budget?”.

Here’s a couple more examples of how to use pre-qualifying filters within your copy:

• “Built for Support Teams with 50+ employees”
• “For Ops Leaders handling 1,000+ tickets a week”
• For “HR Leaders working in The Construction Industry“

This sort of language is self-selecting; it eliminates people outside your ICP and resonates with those who see themselves in the description. You’re trading volume for efficiency, which is exactly what you need when sales capacity and budget are finite especially in high-CPC environments like LinkedIn.

How to run the test: Create two identical campaigns to the same audience. Version A uses broad, benefit-led copy. Version B adds 2–3 crisp qualifiers up front. Measure beyond CTR. Track landing page conversions, demo rates, and post-demo fit.

2) Tofu Landing Page Vs Book Demo Landing Page

Most SaaS ads send cold traffic straight to “Book a demo.” For the tiny group of high-intent buyers (around 5%), that works fine. For everyone else, it’s too much too soon. This test asks: if we warm the audience first with a Tofu educational content landing page, do we generate more qualified demand overall at a lower blended cost?

Tofu: Top of Funnel, typically content that introduced a problem or pain point prior to presenting a potential solution.

The two destinations look like this:

• ToFu landing page: Ungated content such as a blog post “3 Ways Poor Onboarding Creates Hidden Churn.” Short, opinionated, ending with a signpost to MoFu (middle of funnel) content like a scorecard, case study, or webinar.
• Demo landing page: Standard sales page with a product pitch, social proof, and a booking form.

How to run the test: Split your cold audience evenly between the two. Keep the creative, budget, and targeting identical, so there’s only a single variable…the landing page. Then track the full funnel not just initial conversion but also newsletter sign-ups, retargeting engagement, demo requests over 30 days, and pipeline quality.

What you’ll usually see: CTR and engagement are higher on ToFu pages. Your retargeting pools fill quicker, and the demos you book a few weeks later are with people who are usually better informed and further along in their decision-making process. Keep the demo path open for the small percent ready now but scale the ToFu path as your main cold motion.

3) Message And Hook Test

I know…this sounds obvious, but it’s still the most overlooked split test. SaaS teams will A/B test button colours before they ever test the actual message. Yet it’s the hook that determines whether your ICP stops scrolling or not.

Let’s say you were running a campaign for a customer support automation tool, you might choose to test:

• Pain vs aspiration: “Stop drowning in support tickets” vs “Scale customer support without scaling headcount.”
• Old way vs new way: “Manual triage is costing you hours every week” vs “Automated routing clears 60% of tickets instantly.”
• Problem reframe: “Support isn’t your bottleneck …onboarding is.”

How to run the test: Test 2–3 radically different hooks at once. Keep the format consistent (e.g. three single-image ads). Look beyond CTR use scroll depth, video completion, landing page engagement, and demo conversion as tiebreakers.

The winner doesn’t just become your top ad. It becomes the message you propagate across your landing pages, sales deck, and founder LinkedIn posts. Remember what I said about using your Ads as a superfast test lab.

4) Geo Testing

Many SaaS teams’ default to the biggest market, and this usually means the US Then after 3 months they conclude paid ads are too expensive. The truth is, costs, buying behaviour, and competition vary significantly by geographic region. So, testing regions against one another often reveals efficiency gains without needing to change your product or message.

How to run the test: Split campaigns across key regions, for example US vs UK vs Canada. Keep the message and your offer identical, split budget 50/50, and only localise small details like currency and spelling. Measure CPM, CPC, demo rates, and opportunity quality.

Here’s a real-world example, we recently split out a campaign for a SaaS client across UK, US, and Canada and found CPCs were significantly different:

• US: $18.46
• Canada: $14.56
• UK: $12.11

The takeaway here is simple. So long as a geographic region has a large enough audience to support your scaling strategy, there are serious efficiencies to be had using this test. A slightly smaller, cheaper and better-converting market can be a very simple, but very smart play.

5) Retargeting Content Test

Retargeting is where most ad strategies fall apart. Everyone gets shown the same “Book a demo” creative, regardless of their previous customer journey. This can lead to ad fatigue and wasted spend.

The smarter option is to test retargeting content by behaviour. So, what do I mean by that…well, if someone reads a blog about support tickets, retarget them with a case study proving ticket reduction. Maybe someone visited the pricing page, and shows high intent, well retarget them with a 30-day free trial ad. Meet the customer where they are in their journey.

How to run the test: Build separate retargeting ad sets for each of your content types, case study, webinar, scorecard etc. Cap frequency, rotate creatives, and exclude those who have already taken that step. Measure not just CTR but the next action taken.

The impact is two-fold. First, prospects feel understood rather than hammered with stuff they’re not feeling. Second, sales get context, because when a demo is booked, you know which content triggered it, so you can tailor the conversation accordingly.

Putting It All Together

You don’t need a huge team to run these tests, but you will need discipline and cadence. I would start with one or two at a time, document the hypothesis and metrics, and review results weekly.

Ring-fence a portion of budget for testing so it doesn’t get swallowed by short-term campaign pressure. It’s also worth sharing what you learn with your sales team…if they say meeting quality has improved, even though demo numbers are lower, that’s a win.

Conclusion

Ads shouldn’t be treated as a slot machine; they should be appreciated as your fastest feedback loop. Remember, pre-qualifying copy protects your budget in expensive markets. ToFu vs demo destination tests show how your buyers want to progress. Hook tests define the story your company should tell. Geo tests uncover pockets of efficiency you would otherwise miss. Retargeting tests turn anonymous readers into educated prospects who hopefully already trust you.

Run these Ad tests with purpose, log what you learn, and roll out the winners out. That is how SaaS companies can transform paid ads from a cost centre into a test bed, and compounding growth engine.