Amazon Is Buying Cheap Google Clicks. Your Brand Can Too
Published
November 16, 2026
Updated
November 16, 2026
Amazon agencies have leaned harder than ever into Amazon's ad platform, and with demand rising and competition climbing, returns are thinning. On an episode of MerchantSpring Marketplace Masters, Ryan Garrow, Director of Client Solutions at Logical Position and a veteran cross-channel advertiser, argued that the biggest blind spot is tunnel vision: agencies obsess over what they can buy inside Amazon while cheaper, high-intent traffic sits underused on Google and beyond. His core insight is an arbitrage that Amazon itself has been running for years.
The Economics of the Arbitrage
Think of a search funnel, Garrow says, built from the bottom up: people searching for your brand, then people searching for the products you sell but who don't know you yet, then people who merely look interested based on the signals every platform collects. The game is comparing the true cost to acquire a customer across those layers. A $2.50 Amazon click converting at 10% costs $25 to acquire; a 50-cent Google click converting at 2% costs less. Same customer, cheaper route.
Garrow found this by accident while running his own Amazon brand in the organic fertiliser space (so his team could experiment on his money rather than a client's). As a large Google agency, he could see both sides: Amazon was placing minimum click costs on its own search pages while buying cheap clicks on Google, showing up all over Google's auction insights.
In his category, a general fertiliser term carried a roughly $1.50 minimum CPC on Amazon but could be bought on Google for around 35 cents. So rather than pay Amazon $1.50 once a shopper landed on the search page, he bought the same Google click for about 40 cents and pointed it at his own Amazon brand page, intercepting the traffic Amazon was reselling to itself.
Amazon Left Google Shopping, but the Play Didn't Die
Amazon's arbitrage was buying Google traffic, sending it to Amazon, then charging minimum click costs on its search pages, and it grew more complex over the years. Then, across May to July 2025, Amazon abruptly withdrew from Google Shopping ads entirely, having been one of the largest spenders in the auction, and shopping CPCs briefly collapsed by 25 to 30% (Digiday). Garrow's read is characteristically unsentimental: it wasn't about protecting data; it was money; Amazon simply wasn't getting enough clicks on its product pages to justify the Google spend.
The windfall many expected didn't really materialise at scale. People overestimated how much Amazon had been paying per click, and with more merchants crowding in and US economic headwinds pushing brands to accept lower returns to defend volume, Garrow isn't seeing Google Shopping costs stay down. The important nuance is that his arbitrage was never really about Shopping ads, which are hard to feed into Amazon at scale; it lives in text ads, where Amazon still advertises. If you search a term and see only two of Google's four top text-ad slots filled, one of them Amazon, that's a strong signal there's cheap click cost to capture, because everyone else's ad quality wasn't good enough to rank above the organic results.
The Measurement That Makes It Pay
Two mechanics turn cheap external clicks into real return. The first is the Brand Referral Bonus: if you're brand registered and tag your off-Amazon traffic through Amazon Attribution, Amazon credits you a bonus averaging around 10% of the resulting sales against your referral fees, effectively cutting the roughly 15% you'd otherwise pay toward something closer to 5% (Amazon). That alone changes the math on driving external traffic.
The second is ranking. Amazon's search visibility is driven heavily by sales velocity: the more you sell, the higher you rank, and the badges that follow (Amazon's Choice, top seller) lift conversion further, so the rich get richer. Garrow's counsel to new brands is blunt: don't over-polish the page before driving traffic; Amazon's pages already convert if your keywords are right, so get sales, reviews and badges quickly, even at a small early loss.
The catch is that this return is hard to attribute precisely; you can't cleanly say how much a Google-driven sale contributed to ranking third instead of fifth. But the stakes are enormous, because the gap between positions is a cliff: in many categories the top product might do 20,000 units a month while the fifth does 500. Third-party tools that surface sales-velocity data help you see where you actually stand (MerchantSpring's Amazon advertising analytics bring the ad and performance side into one view).
Which Products and Keywords Are Worth It
Not every ASIN is a good external-traffic candidate, and Garrow's filter starts with the competition on Google, not on Amazon. If your Amazon rivals also run aggressive DTC sites bidding on the same Google terms, the arbitrage shrinks. But if your category is full of, say, Chinese sellers who never built DTC sites or advertised on Google, the clicks there can be cheap, and that's where he'd look first. Anyone selling on Amazon should advertise on Amazon; the question is where the additional low-competition clicks are.
Intent is the other half. Garrow thinks in two buckets: brand searchers, who already know you and convert well (keep the branded-defence ad up; it's cheap and it lets you measure brand growth), and non-brand product searchers, who don't. For his fertiliser brand, non-brand terms converted poorly on his own new site against established incumbents, so he sent that non-brand traffic to Amazon, where it held a remarkable 20 to 30% conversion rate he couldn't come close to on his own website. The reason is trust: a shopper will try an unfamiliar brand on Amazon because returns feel safe, and a brand-registered Amazon store page carries credibility even with fewer reviews. That trust transfer is exactly why, when your review count is still thin against thousand-review competitors, routing outside traffic to your Amazon page can beat sending it to your own store.
Where to Send the Click
On landing pages, Garrow's default is to remove friction, not add it. When a Google keyword shows clear purchase intent, sending the shopper to an interstitial blog or a pre-sell page before they can buy just leaks conversions, and the extra tracking data isn't worth the drop-off. The one variation he likes is Buy with Prime. You can run a second, Buy with Prime SKU that a shopper reaches only from off-site, build a Google feed around it, and fulfil from Amazon inventory, which is a neat way to move aging Amazon stock without paying more warehousing fees, while still giving the buyer Amazon's fulfilment and trust. As Buy with Prime matures, he expects its perks to make that trade-off harder to ignore.
The Real Edge: Test Constantly, Across Channels
The through-line is a refusal to treat any channel as a silo. Amazon is too large a share of e-commerce to operate in a vacuum, and TikTok, influencers and content partners all feed it now, so an agency fixated on a single platform is leaving money on the table. Garrow points to under-used arbitrage on Microsoft Ads, where Amazon still advertises and few agencies compete, and expects the next frontier to be AI: as ChatGPT and others build advertising models, bots buying from bots, he predicts more deals like the emerging ChatGPT-Shopify tie-ups, with Amazon likely to follow. None of it is settled, which is precisely why he tells agencies to keep testing.
His practical advice for a stretched ten-to-twenty-person agency is to co-invest with a few trusted clients. Identify where your existing skills transfer, make sure you have the bandwidth, then tell a client you'll test something new at a fair, below-market rate while you both learn, exactly how his own Google-to-Amazon arbitrage started, as a couple-thousand-dollar experiment on a willing client after he'd proven it on himself. Business owners are stuck in their businesses and rarely see where marketing is moving, so the agencies that earn trust are the ones thinking outside the box on their behalf. For agencies ready to push past single-channel advertising, MerchantSpring's guide on leveraging DSP and AMC for agency success is a useful next step.
Compare Your Ad Efficiency Across Every Channel
Arbitrage only works if you can see cost and conversion side by side across channels, which is the whole point of Garrow's approach. MerchantSpring brings advertising and profitability together across Amazon and your other marketplaces and retail-media channels, so you can spot where a click is cheaper, where the return is real, and where your spend is quietly underperforming. See how agencies use MerchantSpring, or book a walkthrough of your own accounts.
Vendor and seller data in one layer
Ordered and shipped revenue, chargebacks, advertising and profitability sit in one data layer, normalised across every channel a brand sells on.
One data layer across every channel you sell on
Every marketplace reports sales, fees and payouts in its own shape. MerchantSpring normalises all of it into one data layer you reach through dashboards, the API, or your own AI tools. The walkthrough takes 15 minutes on a live demo account.
Continue Reading
All articlesWhy AVNs Require New Strategies to Protect Margins in 2027 and Beyond
Webinar Overview Amazon isn't a strategic retail partner. It's a transactional trading relationship, and ...
No Single Channel Should Own More Than 20% of Your Revenue
Most marketplace advice is about getting more out of Amazon. Far less is said about what happens when Amazon ...
Profitability Is Now the Agency's Job, Not Just the Brand's
With Amazon fees climbing, new charges appearing and tariffs squeezing P&Ls, margins for many brands are ...
MerchantSpring is the leading marketplace data platform for Enterprise Brands, Agencies and Developers - powering dashboards, APIs, and AI across 120+ channels.