SaaS Video Ads That Convert: What Actually Makes Them Work

Prefer watching over reading? This whole breakdown started as an episode — watch it below. The article is here if you'd rather skim, save the links, or share it with your team.

A client came to me with a SaaS product. He didn't ask "how much does this cost." He asked me nine questions — real questions about what actually makes a video ad convert.

The questions were so good, I said: let's turn this into an episode. And then into this article.

One hint before we start: question four is the one most SaaS marketers get wrong. It sounds simple. It isn't.

We'll go through real ads — Ramp, Slack, Monday.com, CrowdStrike, Databricks, OpenAI, Notion — and pull out what makes each one work. This is the kind of thinking we bring to every B2B and SaaS video project at Lava Media.

Whether you're planning one-off SaaS video ads for cold traffic or a full series of SaaS marketing videos for the whole funnel, the same rules decide what actually converts.

Question 1: How do you grab attention in the first 3 seconds?

To grab attention, you need a hook — something bright, something sharp. A phrase, or an action in the story, that pulls the viewer in.

A hook can be fake — something loud and flashy just to catch the eye. Or it can hit a real feeling, a real problem your viewer has. The second type works best. When you hit someone's real problem, they think: "Wait, these guys get me." And they keep watching.

To do that, you need to know one thing: what problem does your SaaS solve, and for whom? The same product solves different problems for different audiences. Take Zapier. It solves automation. For a furniture maker, that means automating orders and machines. For a dog shelter, it means automating feeding time. So first you pick who you're talking to. The hook comes after.

One way to hook someone is to call out the problem right away: "You forgot to feed a hundred dogs, because Max was hiding in the corner and you thought he wasn't hungry." Then: Zapier feeds him on schedule and sends you a report on how much he ate. Any shelter owner feels that in three seconds. We never even said "we automate things." We just said the problem out loud.

The second way is metaphor — take the problem and blow it up. Ramp's Super Bowl ad with actor Brian Baumgartner is a great example. The hook is the actor himself — "wait, isn't that the guy who played an accountant on TV?" — and now he's buried in paperwork again. Exaggeration, a familiar face, something interesting on screen.

Even the camera can be the hook. In the Ramp ad, the first two seconds are a hard push-in on the actor's face. Next shot: someone tells him he has a meeting in five minutes. The viewer instantly knows he won't make it. In six seconds they gave you the hero, the problem, a sharp camera move, a cut, and the stakes.

Question 2: How do you show a SaaS interface without putting people to sleep?

This depends on your ad type and where it sits in the funnel.

The Ramp ad sits at the top of the funnel. It doesn't need to show the interface at all — a receipt gets scanned, it goes into the system. No app screens, no buttons. Just the big benefit, shown big, then the Ramp logo, over and over, so the name sticks.

At the top of the funnel, your audience doesn't know you yet. Your job is simple: hammer in the problem, and hammer in that you solve it. Look at CrowdStrike's ad. The interface inside it is barely real — just a rough sketch of a woman "scanning" a Trojan horse. At this stage the message is: "Worried about security? We handle it." What it looks like inside doesn't matter yet.

That Trojan-horse approach is especially powerful for complex IT and FinTech & cybersecurity products, where the actual product is abstract and hard to film.

The middle of the funnel is different. Here people already know you. Say you're Slack and you want them to try a new bot that lives inside the chat. Slack showed this with 3D graphics: the bot starts as a baby, grows up, becomes a "professional." For the actual interface they used simple mockups — the bot writes something, sends a table, nothing fancy. Clean, not overloaded.

Here's the core idea: don't just show the interface as it is. Send the message first. The interface comes second. Think about how Apple sells the iPhone — great photos and the feeling of using the camera, not megapixels and lens specs. Even when you do show the interface, make it clean, simple, animated. Cut the extra tabs and clutter so it supports the message instead of stealing attention.

This is exactly the kind of work we do with animated explainer video. We recently built a custom animated explainer for a B2B SaaS sales team, where the whole challenge was to simplify the product without losing what makes it valuable.

See the full B2B SaaS explainer case study →

That's the same reason the best SaaS demo videos never dump the whole UI on screen: a strong SaaS product demo video shows one benefit clearly and cuts everything else.

Question 3: What converts better — animation, live actors, or screen recordings?

Look at your message first. Then add one more thing: your competitors.

Say you're the first company to build a real work chat, to kill messy email threads — like Slack a few years back. Then a simple, direct ad works fine: "Email is old. It's messy. Here's Slack — a real chat, just for work, with channels, clean and simple." Animated, straightforward. People get it, because you're first.

But if you're Slack today — three years in, a pile of competitors, everyone fighting over who has the smoother button — the competition moves into marketing and creative. You win with a bigger budget, or with better creative. That's why Slack did a musical-style ad that made jokes about itself — office humor, like "let's schedule a meeting to schedule more meetings." The whole format does the work.

So look at what your competitors are doing and how close you are to them. If three competitors are all pushing the same message to the same crowd, repeating the problem won't help. Now it's worth doing something creative — like a dog "talking" about how much better its life got since the shelter started using Zapier.

Here's a real example of that exact challenge — visualizing something you can't easily film. We built an AI and 3D animation piece for a client called AMPD, where the whole problem was showing an abstract decision-making process inside their software. No camera can film that.

See the full AMPD AI + 3D case study →

Question 4: Where does the call to action go — only at the end, or built through the video?

Let's be clear on what a classic CTA means: directly pushing someone to visit a site, install the app, or use a discount. That kind of CTA belongs at the end. It shouldn't break the story — it comes after you've told your story, like a final punch.

But don't forget branding. Your SaaS brand is your URL. Your URL is already a CTA. The rule: show your brand by second five or six. Ramp does this at second six — the logo appears, big. By then the hook has worked, the problem is clear, and the brand becomes the answer. That connection — problem, then brand — is your CTA in SaaS.

Monday.com's ad works the same way. Second one: the hook — something cracks, something falls. Second four or five: the metaphor lands — it's painful when projects fall apart. Second five or six: "Monday.com." Nobody says "go to the website," but everyone understands that's the address.

Short answer: a hard offer goes at the end, but your brand needs to show up by second five or six — and in SaaS, that brand moment is your CTA.

Question 5: One finished video, or a set for testing? Facebook or LinkedIn?

Start with placement — you need to know this first.

If you're going for something like a Super Bowl spot, do what Ramp did: instantly recognizable, a problem everyone understands, a story simple enough that a random guy could explain it to another random guy at a bar.

If you're going into LinkedIn with narrow targeting on a CFO who decides whether to roll your app out company-wide — this is not a creative ad anymore. Maybe 100 to 200 people from your target audience will ever see it. It's a different approach entirely.

You're moving down the funnel: people already heard of you, but now you need to remove real objections. You need a testimonial from a current client — someone's CEO, in a nice office, talking numbers, mixed with real footage, data, and graphics. Take the Databricks ad, where Suntory's CEO talks about their experience. That's the register you need for a CFO.

Or say you want to convince technical integrators to install your app. Now it's about the interface. Show a CTO exactly how the system works inside — how fast a task gets done, two clicks instead of slow spreadsheets passed between departments. Here you need the interface. No creative story — just the facts.

For that audience, a focused SaaS product demo video — upload here, done there, two clicks — beats any creative story.

On hooks — you don't always need multiple versions. With something like the Monday.com metaphor there's not much room to experiment. Or you build several creative entry points at once — a project falls apart one way, a window cracks, a desk breaks. Same metaphor, different visual. It depends on your budget and how many audiences you want to reach.

Question 6: Different videos for different funnel stages?

Yes. Not different versions of one video — completely different videos.

Top of the funnel — you state the core problem and that you solve it. You can do this bright, with a metaphor, or by exaggerating a bad experience with competitors. Take Workbooks CRM: they exaggerate how CRM tools keep upselling you until you can't even use the basic features. Sharp, memorable, and almost everyone who's used a CRM gets it instantly. They barely show the interface.

Middle of the funnel — you can use light tutorials, like Notion does: "click here, click here, and here's your result."

You can announce new features, since people already know you.

You can also use testimonials here — not the heavy, serious kind from the bottom of the funnel, just simpler, more relatable ones. Notion has a great example: a pizza shop using Notion to run their business — building a menu, a schedule, staff training. A classic testimonial, but instead of a bank and a top executive, it's a business anyone can picture.

Bottom of the funnel — no more jokes. Either you go deep on one specific feature, or you talk numbers, straight up. When you're selling hard B2B to a whole company, you hit the CTO with tutorials and demos, the CFO with numbers, the CMO with something else again. Same company, three different conversations: tell the CFO you're cheaper, tell the CTO everything is automated with no engineering headaches, tell the CMO their data will arrive faster and cleaner.

At every stage, this isn't one video reused. These are completely different videos with completely different strategies — the reason a full-funnel plan matters more than any single clip, especially for startups and tech companies building awareness and conversion at the same time.

Question 7: Focus on features and functions, or on benefits and emotion?

It depends on the ad type and the funnel stage.

In a creative ad you can exaggerate the emotion a little — like Monday.com's spot where a woman is so happy she falls off her chair. That level of exaggeration works there.

If your creative ad is calm and simple — like the Trojan-horse idea from CrowdStrike — there's no real user shown at all. It's a pure creative concept, and it doesn't need a "happy customer" moment. The idea itself carries the whole message. So features vs. benefits vs. emotion depends entirely on what kind of ad you're making.

Want this built for your SaaS?

A hook, a clean CTA, the right platform, the right video for each funnel stage — that's the system. If you want it built for your product, that's exactly what we do.

Book a video strategy call with Lava Media →

See you in the next breakdown.