If you’re A/B testing button colors while the first three seconds of your video are flat, you’re testing the wrong thing.
I learned that a long time ago. I started filming skateboard videos when I was 14. Different platform. Same human behavior. If the opening doesn’t make people curious, people move on.
That is even more brutal in B2B SaaS now. Your buyer is tired. Their feed is crowded. They’ve already seen a pile of safe explainers that all feel the same. The standard corporate explainer is boring as hell.
I’m a filmmaker first, entrepreneur second. So I care a lot about the hook. The first line. The first image. The first sound. The first three to five seconds. That’s the part that earns the rest of the ad.
And video is only getting bigger. IAB reported U.S. digital video ad revenue reached $78.0 billion in 2025. So if you’re spending in this channel, the creative has to pull its weight.
Why the Hook Matters More Than Most Teams Think
A lot of demand gen teams obsess over targeting. Or they obsess over the CTA. Sure, those things matter. But the asset matters a lot more than people give it credit for.
Nielsen Catalina Solutions found creative drove 56% of digital sales impact in the campaigns they studied. That lines up with what I see all the time. You can have a good audience and a decent media plan, but if the opening is weak, the ad is dead.
There’s another reason I care so much about memory. Most B2B buyers are not ready to buy today. LinkedIn’s B2B Institute says 95% of B2B buyers are out of market. Then when they do start buying, Gartner found each supplier may get only about 5% of the buyer’s time if three vendors are being considered.
So yes, I care about leads and demos. Of course I do. But I also care about recall. If they do not remember your brand, sales has nothing to work with later.
Humor helps there too. Kantar found humorous ads are more expressive, more involving, and more distinct. I believe that. I’ve seen it. People remember funny. People remember an emotion. Whether it’s a mental laugh or an outside laugh, like an LOL, it sticks.
Step 1: Sell a Testing System Internally
If you’re the marketing champion inside the company, don’t walk into the CFO’s office saying you want to make funny videos.
Say you want to run a creative test. Say you want enough variants to learn what works. Say you want one day of filming to feed months of paid media. That lands a lot better.
I usually tell clients the same thing. Buying one ad is a gamble. Buying a batch gives you a real test. At CinemAds, a package of 10 high-quality commercials is usually around $30,000. That works out to roughly $3,000 per video. That is a lot easier to defend than one expensive hero ad with no backup plan.
This is also why I often write sample scripts and big ideas up front. If you need ammo for your CEO, I want to help you get it. I’ll write until the marketing team and leadership are happy with the direction. Once everyone signs off, I know my writing job is done and my directing job begins.
And one more thing. Don’t confuse impressions with attention. The IAB and MRC say a video can count as viewable when 50% of its pixels are in view for two seconds. That is a pretty low bar. You want your hook doing much more than that.
Step 2: Start with the Pain, the Lingo, and the One Big Message
Before I write hooks, I want the product explained to me in very simple terms.
I say this to clients all the time. Tell me what your product does like I’m 10. Or tell me like I’m 20. Or give me a metaphor for how it works. I’m not a professional in their industry. I’m a professional in my industry. I need the truth from the client so I can turn it into something watchable.
I also want to know who we’re talking to, what pain they already feel, what words they use, and what makes the brand different. That is where the hook starts. Not with the camera. Not with the edit. With the pain point and the audience.
A lot of my best ideas come from that. For HockeyStack, we helped bring in the sales and marketing tension that their audience already understood. That led to the hockey referee idea and the “high dissing” joke. For AskNicely, customer feedback became a host in the park asking strangers to rate their real-life experience. For Aligned, sales software became a dating scenario, with a guy on a date asking about “next steps” and pulling out a PowerPoint.
That’s how I think. I take something dry, find the human angle, and simplify it. Creativity is limitless. It’s infinite. But it has to start from the right message.
Step 3: Write Hooks from Safe to Bold to Unhinged
This is where most teams get too cautious.
They write one script. They polish it to death. Then they call that testing. I don’t do that. I like to write a range.
Some hooks can be safer. Some can be bolder. Sometimes one or two can be a little more unhinged. If leadership wants to stay conservative, fine. I’ll still usually try to get one or two riskier concepts into the batch so you have a real upside play.
I don’t really care for safe videos, but I understand why companies ask for them. So I mix them in. Then I let the ad account tell us what the audience actually wants.
Sometimes the hook is a line. Lavender Joe saying “Do you have 15 minutes to chat?” or “Do you want to explore synergies?” worked because salespeople instantly got the joke. It spoke directly to their world.
Sometimes the hook is pure visual. I directed a telehealth ad for Maple where the opening image was a seemingly naked guy covering himself with an open laptop while romantic music played. Then the record scratched and you realized he was just talking to a doctor. That visual did the heavy lifting right away.
Sometimes I delay the product reveal on purpose. In the Right-Hand Cybersecurity videos, I avoided mentioning cybersecurity at the beginning. I wanted the audience invested in the funny scenario first. Then the brand lands at the end and the message clicks.
I’ve also used bold visual gags just to force curiosity. A tug-of-war with a rope coming out of a computer screen is a good example. Some gag or shock value image makes people stop and say, what the heck is going on here?
The data backs this up too. Google found memorable imagery in the first five seconds is tied to longer view times, and using two or more shots in the first five seconds helps ad recall and consideration. Also, many in-stream ads can be skipped after five seconds. So if your opening is sleepy, you are paying to be ignored.
I keep one hard boundary here. I never want humor that punches down or makes fun of people. You can be bold. You can be weird. You can be over the top. You do not need to be cruel.
Step 4: Turn Five Scripts Into 15 Tests
Once the scripts are approved, I start multiplying the assets.
A very common setup for me is five core 30-second scripts. Then I change the first line of a script several times. One script gets five outputs. The body stays mostly the same, but the hook changes. That gives you a clean A/B test on the opening.
This matters because we just do not know which hook will win until it goes into paid media. I might have a favorite. The client might have a favorite. That doesn’t mean the market agrees.
Then I cut those 30-second spots down into shorter versions. I also deliver each one in horizontal, square, and vertical. So when we’re delivering 15 videos to our clients, they’re actually getting 45 videos.
I still prefer 16:9 because I’m a filmmaker and I love cinematic framing. Our eyes are built more for horizontal. But I live in the real world, and your campaigns need options. Google says campaigns that include vertical videos have seen 10 – 20% more conversions. So I shoot for the best look, then I adapt it for the feed.
This is one reason I push back when a company says they only want one commercial. Shooting a full batch in one day is usually only marginally more expensive, and now you actually have something to test.
Step 5: Shoot Fast, but Keep It Cinematic
High volume only works if the prep is tight.
A normal campaign for me has 30 to 40 days of pre-production. Scripts, revisions, casting, wardrobe, props, storyboards, shot lists, Slack messages, all of it. By the time I get to set, I’ve already visualized the ads in my head.
Then on the shoot day, I keep us in one location when I want volume. A house. An office. A studio. A park. That is how I can move quickly and still keep quality high.
PetMeds is a good example of the volume model. We created 20 assets in two days. AskNicely got a full 10-video campaign from one concept. On LivePerson, we used different backdrops and characters in the same studio world so the campaign felt fresh instead of repetitive.
And through all of that, I’m still following the ABC rule. Always Be Cinematic. I want the ad to look like a mini movie. A commercial that looks like a movie, like a CinemAd.
For Replicant, I shot a rainy diner scene at night with a gritty feel inspired by Breaking Bad and Better Call Saul. For Aligned, I used a dim, moody restaurant because I knew that look would help stop the scroll. For Lavender, I leaned into moody lighting that made people question which platform they were even on.
I also care about sound. A dead opening with no audio is one of the worst ways to start. Google’s YouTube creative guidance says 95% of video is watched with sound on and 70% of watch time happens on mobile. So I want the visual to hit fast, and I want the sound design, music, or first line to help it land harder.
Step 6: Launch a Clean Split Test
If you are testing the hook, keep everything else as clean as you can.
Keep the audience the same. Keep the offer the same. Keep the CTA and landing page the same. Otherwise you’re testing a mess and calling it insight.
Also, give the ads enough budget to breathe. I’ve had clients love the videos, then run them with tiny spend, then wonder why the performance was weak. If the ad budget is too low, that tells you very little. I just handle the bucket of video. The media buying still matters.
When the campaign launches, I’m looking first at CTR and watch behavior. Then I look at view-through rate, CPL, CAC, and whether the leads are turning into MQLs and SQLs. If the hook is doing its job, more of the right people should keep watching and click through.
And the front half matters a lot. In one large study of video advertising, about one-third of ad abandoners left in the first quarter and about two-thirds were gone by the halfway point. That is exactly why I fight so hard for the opening.
If leadership is skeptical about a narrative hook, I’ll often make two cuts. One stays less dashboardy and ends with the tagline. The other adds a quick solution shot or a UI moment at the back. Then you can run both and compare.
My favorite funnel setup is simple. Use the funny, cinematic brand video at the top. Then retarget people who watched more than 50% with a more traditional explainer that shows the product and answers deeper questions. That is a much better sequence than dumping the dashboard on a cold audience in the first frame.
Step 7: Read the Results Honestly, Then Scale the Winner
When the data comes back, be honest.
The script your team loved in Slack may not win. The one you thought was too weird might be the best performer. Paid media decides. That is why I keep saying to test a range.
I’ve seen the upside when the creative and spend are both there. Sparkpolo ran 10 of our assets and hit a $17.16 cost per lead with a 3.73% CTR. HockeyStack’s cinematic ads drove 10x higher CTR and 3x higher ROI than their previous formats. Replicant recouped its production investment from a single CTV sale and came back for more.
I also pay attention to recall. That matters a lot in B2B.
Dave Lehmkuhl said waiting for the next Lavender video was “like waiting for next episode of your favorite Netflix show.” Another viewer said it was “better than 99% of shows on Netflix.” I’ve had booth workers recognize Lavender Joe at Collision in Toronto. The actor was even recognized in Italy.
Those are not small signals. That tells me the campaign moved out of the ad bucket and into culture, at least for that niche. When a viewer starts remembering the character by name, the hook is doing real work.
Step 8: Turn the Learning Into a Monthly Engine
A/B testing should not be a one-time event.
Once you find something that works, keep the big idea and rotate the execution. Change the first lines. Change the scenarios. Change the pain points. Keep feeding the account fresh versions so you do not burn out the same ad.
That is why I like monthly retainers. Month one gives you data. Month two gets sharper. Month three gets sharper again. We did that with Lavender over a year and produced more than 300 videos for under $120,000 CAD. Same universe. Same character. Fresh hooks.
This is also where I think human production still has a big advantage. I use AI for brainstorming and writing. I think it’s useful. But there’s still that human nuance that AI doesn’t have. A look from an actor. A funny pause. A line read nobody planned. Those moments matter.
And once a hook proves itself on social, I like pushing that thinking into CTV too. A funny or cinematic ad hits much harder on a 65-inch screen. I know that from real life. My wife and I have literally repeated jokes from ads we watched on Netflix because the joke actually landed.
Final Thought
If you’re the person inside the company trying to get this approved, I get it. You need proof. You need assets. You need leadership buy-in. You need something your ad account can actually work with.
So don’t bet everything on one safe video.
Test the opening. Test safe against bold. Test dialogue against visuals. Keep the top of funnel focused on attention and memory. Then retarget with the explainer if you need the deeper product story.
I don’t really care for safe videos, but I do care about learning. And the fastest way to learn is to stop pretending one ad can do everything.
CinemAds exists because brands need funny, memorable, engaging, high-quality video ads. In crowded SaaS markets, it’s the best marketing that wins.
Frequently Asked Questions
How often should we rotate video hooks to prevent ad fatigue?
Rotate them the second your CTR drops or CAC spikes. I usually tell clients to refresh hooks every three to four weeks. That is exactly why we shoot 15 variations in one day. You need a deep arsenal to feed the machine without burning out your best concepts.
Should we use subtitles on our A/B tested video hooks?
Yes, use subtitles, but never ignore audio design. While Google reports 95% of YouTube is watched with sound on, social feeds vary. A killer hook needs strong visuals, punchy captions for mobile scrollers, and cinematic sound to lock in attention once they stop.
Will humorous hooks alienate serious enterprise B2B buyers?
No. Enterprise buyers are still exhausted humans staring at a crowded feed. According to Kantar, humorous ads are 14 points more involving. If the joke is rooted in their actual daily pain – like friction between sales and marketing – they will laugh, remember your brand, and click through.
Why do we need to A/B test vertical video hooks for B2B SaaS?
Because that is where the attention lives. I prefer cinematic 16:9, but Google found vertical assets drive 10 to 20% more conversions on mobile. We always cut horizontal, square, and vertical outputs from the same shoot so you can adapt your best hooks natively.
If the hook is five seconds, how long should the total video ad be?
Keep top-of-funnel ads under 30 seconds. A large UMass study found two-thirds of ad abandoners leave by the halfway point. Hit them hard in the first three seconds, deliver the joke, tease the SaaS solution, and get out. Save the 90-second deep dive explainers for your retargeting.
