Blog: CTV Just Became a Performance Channel for the Mid-Market. Most Brands Haven’t Noticed Yet.

The economics of CTV haven’t fundamentally changed. What changed is the measurement infrastructure around it. Tools that were once enterprise-only—like incrementality testing, marketing mix modeling (MMM), brand lift studies, and household-level attribution—are now accessible to brands spending a fraction of what enterprise advertisers spend. That shift turned CTV from an awareness channel into a performance channel for anyone willing to measure it properly.

Most mid-market brands haven’t caught up. But their largest competitors have.

The attribution problem that hides CTV’s real performance

When a mid-market brand tested CTV three years ago and saw flat results in their conversion reports, the conclusion was almost always: CTV doesn’t drive revenue for us.

The actual conclusion should have been: Our measurement model can’t see what CTV is doing.

Last-click attribution—still the default in most marketing dashboards—assigns 100% of conversion credit to the final ad or link a customer touched before purchasing. It’s clean, it’s easy to report on, and it’s wrong: Last-click attribution systematically under-credits any channel that influences consumers earlier in the journey. Which is most of them. And especially CTV.

The way I think about it: imagine you spend a night out with friends. A glass of wine at the first place. A margarita at the next spot. Champagne at a wine bar. And then you end the night at a pub with a Miller Lite. The next morning, you wake up hungover and tell yourself, “That last beer got me drunk.”

Through the lens of last click measurement, that last beer did get you drunk. Your CRM will confirm it. You and I both know it wasn’t the beer.

CTV is one of those earlier drinks. It almost never gets the final click. Someone sees a CTV ad on Tuesday, searches your brand on Friday, clicks a Google ad, and converts. Last-click hands the win to Google. CTV gets nothing. Multiply that across thousands of conversions, and the channel looks dead in your reporting, even when it’s the reason the search happened.

The shift that matters isn’t a new ad format or a better targeting capability. It’s that mid-market brands now have access to measurement tools that can accurately see past the last click.

How do you measure CTV advertising?

Four tools are doing most of the work and none of them require enterprise budgets anymore.

• Incrementality testing. Split a comparable audience into two groups. One sees your CTV ads. The other doesn’t. Compare conversion rates. The difference is the causal lift your CTV spend produced (i.e. conversions that wouldn’t have happened without the ad). This is the gold standard, and the bar to run a clean test has dropped dramatically. Several DSPs and measurement vendors now support it natively. (We’ve written a deeper primer on how incrementality testing actually works if you want to go further.)

• Marketing mix modeling (MMM). A statistical model that looks at your full marketing mix — every channel, every spend level, every external variable like seasonality or promotions — and estimates how much revenue each channel contributed. MMM used to require six-figure engagements with specialty consultancies. The current generation of MMM platforms has dropped the cost and complexity dramatically, and the model is built to credit channels like CTV that influence revenue without owning the final click. Where incrementality answers “did this specific test produce lift?”, MMM answers “across our entire mix, where is each dollar actually earning its keep?”

• Brand lift studies. Survey-based measurement that asks the simple question: did your CTV campaign change what people think and feel about the brand? Ad recall, brand awareness, purchase intent. Platforms like StackAdapt have this built in. It’s not a substitute for incrementality, but it’s a useful read on whether the creative is doing its job.

• Household-level attribution. Connects the TV that served the ad to the phone or laptop that later visited the site, using shared IP addresses and device graphs. It’s directional, not precise — the assumption that everyone on the home WiFi is the same person breaks down quickly — but it gives you a real signal where you used to have none.

You don’t need all four measurement tools. But you do need at least one, and you need to run it before you write CTV off.

What CTV performance looks like when it’s working

Here’s a pattern worth understanding. One of our clients runs a brick-and-mortar social experience venue. Strong brand, loyal customer base, two years of compounding YoY growth from SEO and paid media that we’ve built and scaled together. The fundamentals of the program are working.

In Q1 of this year, growth stalled. Direct traffic (i.e. returning customers and brand searchers) dropped 9% in sessions and 13% in conversion rate compared to the same period last year. Macro pressure on discretionary spend explains some of it. But the brand search trend pointed to something more structural: fewer people were searching for the brand at all. The top of the funnel was shrinking.

Our competitive analysis showed their largest competitor is doing something about that. This competitor has 400+ locations across three concepts, $100M–$140M in annual revenue, and — here’s the relevant part — is running geographically targeted CTV campaigns by ZIP code. 

The competitor’s head of marketing called CTV a top five revenue-driving advertising source in interviews. And the company is scaling the CTV strategy across all three brands.

Their reported results: 187% blended ROAS above goal. 153% lift in conversion rate.

This doesn’t mean that existing channels like Google have stopped working for the client. But Google Search can’t grow an audience that’s getting smaller; they can only convert the demand that already exists. Closing that gap with proportionally-matched funnel expansion on CTV is a top priority.

The real question: where does the next dollar go?

Most mid-market marketers reading this aren’t deciding between CTV and nothing. They’re deciding between adding more spend to the channels that are already working — Google, Meta, the trackable performance stack — and putting that same spend into a channel like CTV that fills a different part of the funnel.

For a long time, the trackable channels won that argument by default. The reporting was cleaner. The attribution was tidier. The ROAS numbers looked better in the deck.

But every channel hits a point of diminishing returns, and on the trackable channels that point often arrives silently. Your dashboard keeps reporting conversions. Your ROAS holds. But a growing share of those conversions are people who would have bought from you anyway: brand searchers, returning customers, retargeted visitors already deep in the funnel. You’re paying to take credit for revenue you would have gotten for free. The click-based dashboard can’t tell you that, because it wasn’t built to.

That’s the moment when the next dollar stops belonging in Google and starts belonging in a channel that’s actually expanding the audience. CTV is one of the most credible candidates for mid-market brands right now — addressable, measurable, and proven in market by the enterprise versions of the same business models. The brands that move on it early get to grow the top of the funnel before their competitors close the window.

If you don’t know which side of that diminishing returns line your current spend is on, that’s where to focus. Run an incrementality test on the channels you’re already investing in. The answer will tell you where the next dollar should go.