CTV Video Ads: Winning in 2026 with $750K Campaigns

The video ad field on Connected TV (CTV) is a tough place to be in October 2026. A recent Statista report shows CTV devices are in about 90% of US households, so simply being there isn’t enough. If your campaigns are going to work, you need to keep up with the latest video ad trends, because brands need a real strategy to get attention and actually drive conversions in this packed environment.

We just wrapped up a big CTV advertising campaign for “Home Harmony,” a direct-to-consumer (DTC) brand that sells smart home gear. Their goal was simple: get the word out and sell more of their new AI-powered thermostat. This wasn’t some small test run. We went in with a $750,000 budget over a tight six-week duration, from early September to mid-October 2026, because we needed hard data to guide our next moves.

Our plan wasn’t just to spray and pray with a 30-second spot. We used a mix of broad reach and extremely specific targeting. You can’t just treat CTV like old-school broadcast TV. The audience is used to premium content and won’t tolerate lazy ads. We wanted the ad to feel like a useful suggestion, not an annoying interruption, which meant we had to nail the creative optimization and lean heavily on our data segments.

Campaign Strategy: Beyond Basic Demographics

For Home Harmony, we started with a core audience of homeowners and tech enthusiasts aged 25-54 with a household income over $80,000, but we went much deeper than that. We layered in third-party purchase intent data for smart home products from sources like Nielsen and combined it with Home Harmony’s own first-party CRM data. This let us build out some powerful custom audience segments on platforms like Google Ad Manager and The Trade Desk, specifically targeting households that had already shown interest in energy-efficient tech or were subscribed to home automation newsletters.

We also added a geographic layer, concentrating our ad buys in suburban areas with a high density of single-family homes. For instance, in the Atlanta metro area, we didn’t just buy “Atlanta”, we focused on specific zip codes like 30342 (Sandy Springs) and 30076 (Roswell) known for those homeowner demographics. That level of focus cut down on wasted ad spend. We also stuck to dayparting, running most of our ads during evening prime time (6 PM to 10 PM) and on weekends, which is when we knew our audience was actually sitting down to watch.

Creative Approach: Dynamic Storytelling and Interactivity

The creative itself was everything. We produced three different 15-second video ads, with each one highlighting a specific thermostat benefit: energy savings, comfort, or ease of use. Instead of just running one static ad, we went with a dynamic creative optimization (DCO) setup. This personalization, though a pain to configure, was a huge advantage. It meant we could change elements inside the ad, like the exact energy saving stat or the call to action, based on the viewer’s location or other data we had. A viewer in a cold state might see an ad focused on heating efficiency, while someone in Florida would get messaging about cooling.

The real hook was a scannable QR code that we placed on-screen for the final five seconds of the ad. When someone scanned it, they were taken directly to a dedicated landing page about the thermostat. This gave us a direct path from the couch to the website, bypassing the clunky “see an ad and search for it later” behavior. We did test a version with a simple URL overlay, but the QR code beat it hands-down on direct click-throughs.

Performance Metrics: A Deep Dive into the Data

The campaign delivered some strong numbers. Across the six weeks, we logged 25 million impressions on platforms including Hulu, Roku, and Samsung TV Plus. The campaign’s overall Click-Through Rate (CTR), driven almost entirely by QR code scans and some direct URL entries, came in at 0.72%. That’s a good result, sitting above the 0.6% industry average for non-skippable CTV ads reported in IAB’s 2026 Digital Video Ad Spend Report.

In terms of actual sales, we tracked 6,200 thermostats sold directly from the campaign. This worked out to a Cost Per Conversion (CPC) of $120.97, which we were happy with since our initial target was $150. Since the thermostat sells for $340, the campaign generated a 2.8x Return on Ad Spend (ROAS). A very healthy return.

We also looked at top-of-funnel actions, and our Cost Per Lead (CPL) for website visits that led to an email signup or product page view was just $2.50. This gave us a solid pool of engaged users for future remarketing.

What Worked Well: Precision and Personalization

A few things really made this campaign click. The dynamic creative optimization was a huge success. By tailoring the ad’s message, we saw a 20% higher conversion rate from those personalized ads compared to the generic versions. It really spoke directly to the viewer’s immediate needs, going beyond just listing product features.

The QR code integration was also a big winner. It created a frictionless path from the passive TV viewing experience to an active web session on a phone. We found that a lead from a QR code scan had a 3x higher conversion rate than someone who manually typed in the URL. This just shows how important it is to remove every possible point of friction on CTV.

Our audience targeting which blended first-party customer lists with third-party purchase intent signals, got our ads in front of the most likely buyers. This precision is what directly led to our good CPC, and it proved that smart targeting is the best way to reduce wasted ad spend. We also noticed that households with multiple CTV devices had higher engagement, which probably points to a more digitally native and receptive audience.

What Didn’t Work as Expected: Early Learnings

Not everything was perfect. My initial hypothesis was that a longer, 30-second ad would give us more room for storytelling, but I was wrong. Early A/B tests showed the 15-second spots completely outperformed the 30-second versions on completion rates (92% vs. 78%) and engagement. Viewers in an ad pod just want you to get to the point. Quickly.

We also burned some budget at the start on broad, demographic-only targeting to see what would happen. It was less efficient. While it got us a lot of impressions, the CPL was nearly double ($4.80) compared to our refined segments, and the conversion rate was way lower. This just confirmed that hyper-targeted segments are essential for making CTV work, particularly for a DTC brand with a specific product.

The other ongoing headache is attribution modeling. It’s still incredibly complex to prove the true influence of a CTV ad on a purchase path that might span multiple devices over several days. We used a blended model with view-through conversions and our analytics platform’s cross-device tracking, but getting a definitive, single-source answer is still an industry-wide challenge. We’re still working to refine this, maybe with more advanced household graph integrations on the next campaign.

Optimization Steps Taken: Iteration is Everything

We were actively making changes mid-campaign based on what the data was telling us. Seeing the performance difference, we immediately shifted 15% of the budget from the 30-second ad variants over to the 15-second ones. Within the first two weeks, we also paused the broad demographic segments and funneled that money into our more precise first-party and purchase-intent audiences. That move alone dropped our overall CPL by 15% within a week.

We even A/B tested the QR code’s on-screen placement. We found that shifting it slightly to the left and bumping its size by 10% during the ad’s final five seconds produced a 7% increase in scan rates. These small, data-driven tweaks made a real difference.

Plus, we adjusted our frequency capping. We started at 4 impressions per user per day but found that dropping it to 3 impressions per user per day for our most engaged segments prevented ad fatigue, while we could actually increase it to 5 for colder segments to build awareness. This kind of careful frequency management is good practice for respecting the viewer’s experience on CTV.

The Home Harmony campaign in October 2026 showed that successful CTV advertising today requires a mix of precise targeting, interactive creative, and non-stop, data-driven optimization. Brands have to get out of the old broadcast mentality and truly understand their audience’s habits on these platforms. The future of CTV ads is about creating relevant experiences that produce measurable business results. As you plan your own campaigns, think about how your video ad audio can expand your reach and make sure your message lands, even when people are only half-watching.