StackAdapt's State of Programmatic Advertising 2026 report found that campaigns using dynamic creative optimization deliver a 32% higher click-through rate and a 56% lower cost per click than static ad campaigns. The same report found advertisers using first-party data or AI-based contextual targeting see up to 2x higher return on ad spend than campaigns relying on third-party data. The findings matter for any small business still paying for cold, third-party-targeted reach.
What The Report Found
StackAdapt, a demand-side advertising platform, published the report on June 24 using its own internal platform data rather than an independent third-party survey. The company didn't disclose a sample size, but the numbers line up with a case study in the same report from Vallo Media, an agency that used dynamic creative to re-engage shoppers who had abandoned their cart. That single campaign produced a 60% lift in click-through rate and generated 30% of the client's total ad-attributed revenue while spending just 12% of the overall campaign budget on it.
Dynamic creative optimization automatically swaps images, headlines, and calls to action within a single ad based on who's actually viewing it, instead of running one static version to everyone. Rather than a media buyer manually building ten versions of an ad and guessing which performs best, the system tests combinations in real time and shifts spend toward whichever version a given audience segment actually responds to. Most major ad platforms, Meta and Google included, have offered some version of this for over a year, but adoption still lags well behind availability.
What It Means For Your Ads
For a small business, the 56% cost-per-click number is the one that matters most. Third-party targeting, buying access to an audience someone else built, has gotten more expensive and less accurate since privacy changes limited how much of that data platforms can actually use. First-party data, the emails, purchases, and site visits a business already owns, doesn't have that problem, and it's what the AI-based targeting behind the 2x ROAS number is actually built on.
The practical shift here is less about creative variety and more about what data is feeding the algorithm in the first place. A business running the same three ad images to everyone while optimizing purely on third-party audience data is paying the old tax on both ends: worse targeting and no creative testing. Feeding a system real first-party signals, past customers, cart abandoners, email openers, is what produces results like Vallo Media's.
First-party data doesn't require a data science team to collect. It's the email list from past customers, the phone numbers captured at booking, the people who filled out a form and never converted, the visitors who added something to a cart and left. Most small businesses already have this sitting in a CRM or an email platform, unused for ad targeting because uploading and managing custom audiences manually across every platform is tedious enough that it just doesn't happen.
This is exactly the gap platforms like KenjiAI (kenjiai.com) are built to close for smaller businesses that don't have an in-house data team: pairing first-party signal with automated ad management instead of guessing at a cold audience.
What To Watch
None of these numbers come with a disclosed sample size or independent audit, they're StackAdapt's own platform data plus one client case study, so treat them as directional rather than gospel. The concrete move either way: check whether your current ad account is actually running dynamic creative rather than a single static ad per audience, and check whether your targeting is built primarily on your own customer data or a purchased third-party audience. If it's the latter, that's the lever worth pulling first, before touching budget or creative concepts entirely.
Expect more platforms to push first-party data tools harder over the next few quarters as third-party targeting keeps getting more expensive industry-wide, not less.
Published by the Media Traffics | KenjiAI team. kenjiai.com
