Privacy Sandbox Didn't Save Paid Ad Measurement. Smart Banners Are Why Email Wins the Budget.

In April 2025, Google did the thing it had spent five years promising to do, then refused to do it. Third-party cookies stayed in Chrome. Performance teams exhaled, and the trade press framed it as relief. I read it as the opposite, and so should anyone signing off on where the next performance dollar goes. The cookie reversal solved a problem that was never email’s to begin with, and it left the real wound in paid measurement untouched. That is precisely why email, and the Smart Banners that make its revenue legible, is the channel that wins budget in 2026.

Here is the part nobody wants to say out loud. Keeping cookies in Chrome does nothing for the 60 to 75 percent of iOS conversions that Apple’s App Tracking Transparency already hides from ad platforms. It does nothing to close the revenue gap regulators measured when Google tested its own replacement. The reversal was an admission that the replacement underperformed cookies, not a restoration of clean measurement.

Email never depended on the cookie. Its measurement runs on identity you own. No single platform’s policy decision can grant that or take it away. Once you internalize that difference, the capital-allocation question stops being “when will attribution get fixed” and becomes “why is the larger budget still pointed at the channel where measurement keeps eroding.”

The reversal everyone misread

On April 22, 2025, Anthony Chavez, Google’s VP of Privacy Sandbox, wrote that the company had “made the decision to maintain our current approach to offering users third-party cookie choice in Chrome, and will not be rolling out a new standalone prompt for third-party cookies.” Ad Age called it “a complete reversal of its more than five-year plan to phase out the internet trackers.”

Read the rest of that post and you find the real signal. Google reframed the Privacy Sandbox APIs as having “a different role to play.” Translation: the replacement is being demoted, and the cookie is not being celebrated as a fix. Regulators had already done the math. The UK’s Competition and Markets Authority review in June 2025 found Privacy Sandbox produced roughly 30 percent lower per-impression publisher revenue than cookie-based targeting, with Google’s own tests landing around 27 percent lower. A peer-reviewed PNAS analysis put it more bluntly: removing third-party cookies cuts publisher ad revenue by 29.1 percent, and Privacy Sandbox recovers only 4.2 percent of that loss.

So the cookie stayed, because the alternative tested worse. That is a retreat dressed as a reprieve. It tells you nothing reassuring about the long-term health of platform-granted measurement, and it should change how you think about which channel deserves your marginal dollar.

Two privacy taxes paid media still pays

Paid media carries two separate measurement taxes, and the cookie decision pays down neither.

The first is Apple’s App Tracking Transparency. Adjust’s 2025 data puts industry-wide opt-in around 35 percent, which means roughly two-thirds of iOS users decline tracking and their conversions never reach device-level attribution. Chrome keeping cookies does not touch a single iOS conversion, because that signal loss happens at the device, on a platform Google does not control.

The second is the Privacy Sandbox revenue gap I just described. Even in the world where cookies survive, the trend of platform signal degrading under privacy and regulatory pressure has not reversed. It has only paused.

Stack those on top of the economics paid teams are already living with. Average ecommerce ROAS fell to 2.87 in 2025, down across 13 of 14 industries. Customer acquisition cost is up 40 to 60 percent since 2023, with Shopify’s own data showing merchant CAC climbing from $274 to $318 year over year. Meta CPMs rose about 20 percent, and Google CPCs about 12.88 percent. You are paying more to reach people and seeing less of what happens after the click. For the full picture on where email sits against these numbers, read our 2025 email performance benchmark report.

Why the cookie decision changes nothing for email’s structural advantage

Email’s measurement was never built on the cookie, so the cookie’s survival is irrelevant to it. When someone opens your email and clicks, that action is keyed to an email address you already hold. You do not borrow that identity from Apple or Google. You own it. It is first-party by construction.

This is why a click-to-conversion rate (CTC) measured on email behaves differently from a paid attribution model. ATT opt-out has zero effect on it. Cookie deprecation, had it happened, would have had zero effect on it. Mail Privacy Protection inflates opens, but the click is still real and still attributable. The signal does not pass through a platform that can decide to withhold it. We made the longer version of this case in our piece on the owned media advantage, and in the math behind treating email as a performance channel.

Smart Banners measure on identity you own, not platform-granted signal

This is where Smart Banners do the real work. A Smart Banner is a banner that renders at open time — one in the header of the email and one positioned above the email’s footer — deciding what each subscriber sees based on first-party signal you already have. Because the decision and the click both sit on owned identity, the resulting CTC is privacy-durable. It survives Mail Privacy Protection, it survives ATT, and it would have survived cookie deprecation.

That matters for reporting, not just principle. While paid teams reconcile modeled conversions against a shrinking pool of observed ones, Smart Banners give you a click number that means the same thing this quarter as it did last quarter. When your measurement stack is losing resolution everywhere else, that stability is the asset. You are not modeling a guess. You are counting a click against an identity you control.

The capital-allocation case for reallocating toward email

Now the part that actually decides budgets. For most ecommerce brands, paid media spend is five to ten times the size of the email program’s investment. That asymmetry is the whole opportunity. A reallocation that registers as a rounding error on the ad side lands as a step-function upgrade on the email side.

I am not arguing email replaces paid. Paid acquires the audience; email monetizes and retains it. The argument is narrower and harder to dodge. The marginal performance dollar should flow to the channel where measurement quality compounds in your favor every quarter, not toward the one where it erodes. A brand spending $2 million a month on ads could move two percent of that into email content and engineering, and the email team would feel a budget it has never seen. The ad team would barely notice the line item move. We unpacked the budget mechanics in The $318 Problem, and the first-party data dependency in our look at why your email program is the upstream source for ad targeting.

How Smart Banners turn the privacy-durable thesis into reportable revenue

A thesis a CFO cannot read is just an opinion. Smart Banners turn the privacy-durable argument into numbers that sit next to your CAC and ROAS lines.

Block-level attribution reports CTC and revenue per thousand (RPM) for each use case and variant, on a 7-day click-based window. The Smart Banner positioned above the footer consistently pulls click-attribution revenue out of proportion to its real estate. Across the Zembula platform over the trailing 30 days, personalized content drove about $4.2 million in attributed email revenue across 17 active vendors at a 59.48x average ROAS on roughly 1.92 billion personalized renders. Smart Banner content averages about 13.6 percent CTC against a roughly 2.5 percent baseline for the email as a whole.

Then there is the test paid media structurally cannot run. A collapsed-pixel holdout shows a slice of your audience the block rendered down to nothing, which gives you a true A-versus-nothing incrementality read. Paid platforms cannot offer that, because you cannot withhold an ad auction from yourself and still measure the counterfactual cleanly. Pair that with daily, ad-style reporting, trend lines instead of per-campaign roll-ups, and email starts to look and report like the performance channel it always was. The full mechanics live in our ultimate guide to Smart Banners.

Key takeaways

• The cookie reversal was a retreat, not a rescue. Regulators measured roughly 30 percent lower per-impression revenue from Privacy Sandbox, and a PNAS analysis found it recovers only 4.2 percent of lost cookie revenue. Google kept cookies because the replacement tested worse.
• Two privacy taxes still hit paid, and the cookie decision touches neither. iOS ATT hides roughly two-thirds of iOS conversions, and the Sandbox revenue gap remains.
• Email measurement is owned by construction. Smart Banners — one in the header of the email and one above the footer — decide and measure on first-party email identity, so CTC is privacy-durable through ATT and Mail Privacy Protection.
• Reallocation is asymmetric. A rounding error of ad budget is a step-function upgrade for email, the channel where measurement quality compounds in your favor.
• Smart Banners make it reportable. Block-level CTC and RPM, plus a collapsed-pixel holdout, give finance the A-versus-nothing incrementality read paid media cannot produce.