How PMG Brought in New Customers for IQBAR Using Full-Funnel Advertising

Full analysis

PMG ran video, display, and sponsored ads for IQBAR, a protein bar brand, and growth still stalled. When they looked inside Amazon Marketing Cloud, Amazon's data clean room, they found out why: they were hitting the same buyers over and over. Then they used five years of purchase history to suppress existing customers and chase only new ones. Ari Paparo interviewed PMG's Ryan Walker about it at Amazon unBoxed. That's the useful kernel. The rest is a case study ad for Amazon Ads.

How hard is this to undo? For an operator, nothing here demands a decision. This is intelligence about where retail media is heading, not a fork in the road. Easy to act on, easy to ignore.

What's actually being discussed: Amazon is pitching its walled garden as a single place to buy the whole funnel and measure it against real sales. The case study is the sales pitch.

The council

The Market Analyst Walker said the thing out loud: "a movement towards more centralized buying and I think Amazon's at the forefront of that." Translate that for your P&L. Centralized buying means one platform owns targeting, creative tools, inventory, and the measurement that grades all three. That is budget consolidation, and every dollar that consolidates into Amazon is a dollar that doesn't flow through an independent DSP, an SSP, or a third-party measurement vendor. The Trade Desk makes the identical "central buying layer" claim, so this is a positioning war, not a settled fact. For a generalist: the big platforms want to be the only checkout lane, and they grade their own receipts.

The Skeptic Amazon's closed-loop measurement is closed inside Amazon. It ties a Thursday Night Football impression to a purchase, but only a purchase on Amazon. IQBAR sells in grocery, in club stores, on its own site. None of that shows up. So "fully closed loop" is true and incomplete at the same time. Walker is a paid Amazon partner speaking at Amazon's own conference, and he framed the offer as best-in-class without once naming the off-Amazon blind spot. The reach-saturation diagnosis is genuinely good work. The measurement claim is a walled garden grading its own homework and calling the grade objective.

The Operator The suppression move is what's worth copying, and it costs you nothing. Most brands run "acquisition" campaigns that quietly re-buy their existing customers because the platform optimizes toward whoever converts cheapest, and existing customers always convert cheapest. Suppress five years of buyers and your true cost to acquire a new customer jumps, often ugly. That's not a reason to avoid it. That's the number you should have been looking at all along. Any agency or in-house team with clean-room access can run this diagnosis next week. If you're a measurement vendor, this is the use case to sell: "your acquisition number is fiction, here's the real one."

The Customer / End User (the brand) For a brand that lives on Amazon, this pitch is close to irresistible. One login, full funnel, tied to sales. For a brand with real brick-and-mortar or direct-to-consumer revenue, buying this story whole means optimizing to the slice of sales Amazon can see and flying blind on the rest. You'd over-weight whatever Amazon measures and starve whatever it can't. The protein bar category is heavily club and grocery. So even IQBAR, the hero of this case study, almost certainly has sales Amazon never counted.

Where they part ways

The real disagreement is whether "closed-loop" is a feature or a trap. The Operator loves it: finally a hard acquisition number instead of a brand-lift guess. The Skeptic says the hard number is hard only inside one store, and a precise number of the wrong thing is more dangerous than a rough estimate of the whole thing, because precision makes you trust it.

The second split is on centralized buying. The Market Analyst reads Walker's enthusiasm as a real budget shift already underway. The Skeptic reads it as a partner reciting the host's talking points at the host's event. Both can be true: the trend is real and the messenger is paid.

What it hinges on

It comes down to one question for any brand weighing Amazon's full-funnel pitch: what share of your sales does Amazon actually see? If it's 80%, the closed loop is close enough to the whole picture and you should lean in hard. If it's 30%, you're handing budget allocation to the retailer with the most reach and the narrowest view, and you need an independent measurement layer sitting above Amazon to catch the rest. The council leans toward: steal the suppression diagnosis, resist the single-platform measurement story.

Impact on independent ad-tech is low and indirect from this one episode. No announcement, no data release, no deal. What it confirms is direction. Amazon wants the whole funnel and the scorecard, and it's using agency partners to tell the story for free.

Prediction: Amazon Ads revenue will grow faster than 20% year over year in the quarter reported on Amazon's Q4 2026 earnings call in early February 2027.

Confidence: Medium The full-funnel consolidation pitch is landing, but ad growth is lumpy and comps are tough.

Why: This case study shows the mechanism Amazon is selling at scale: one place to target, create, buy CTV, and measure against real purchases, which pulls both upper-funnel brand budgets and lower-funnel performance budgets onto the same platform. Brands that were running awareness elsewhere and performance on Amazon now have a reason to move the awareness spend too, because Amazon ties the impression to the sale and nobody outside the walled garden can match that on-platform proof. Amazon's ad segment has been printing growth in the mid-to-high twenties for several quarters, well ahead of the broader digital ad market, and the Prime Video plus Thursday Night Football inventory gives it new high-value supply to sell into that same loop. The opposite outcome, growth dropping below 20%, would require either a macro pullback in ad spend or the consolidation story stalling, and nothing in the current demand picture points that way.

Revisit by 2027-02-15: We're right if Amazon reports advertising services revenue up more than 20% year over year on its Q4 2026 earnings call. We're wrong if that growth comes in at 20% or lower.