Published: 11th July 2026
A £20,000 monthly Amazon budget can show an acceptable ACoS while your wider acquisition model is quietly losing money. That usually happens when Amazon, Google and Meta are managed as separate media silos, each chasing its own attributed return. Knowing how to structure hybrid ecommerce ad accounts is not about forcing every channel into one reporting template. It is about giving every pound a clear commercial job, a controlled destination and a measurement framework that reflects profit.
A hybrid brand sells through Amazon and its own site, often while using paid social and search to create demand across both. The account structure must recognise that customers do not behave according to platform reporting. They may discover a product on Instagram, search the brand on Google and buy on Amazon because of Prime delivery, reviews or familiarity. If every channel is asked to take full credit for that sale, your budget decisions become expensive guesswork.
Start with the operating model, not campaign naming conventions. Before building or restructuring anything, decide what each channel is expected to do and where it is allowed to send traffic.
Amazon should usually own the conversion of high-intent marketplace demand. Sponsored Products captures shoppers searching for a product or competitor on Amazon. Sponsored Brands builds brand presence and directs traffic to a Store or suitable product selection. Sponsored Display supports retargeting, product targeting and audience expansion where the economics justify it.
Google Search should protect branded demand, capture broader category intent and support direct-to-consumer revenue where the site can convert profitably. Meta and TikTok are generally better used to create demand, build audiences and retarget engaged users, rather than being judged solely on last-click platform ROAS. The exact split depends on your margin, repeat purchase profile, retail pricing and Amazon’s share of total revenue.
The mistake is treating Amazon as just another performance channel. Amazon is both a marketplace and a conversion environment. It has different shopper intent, different attribution windows and a different relationship with pricing and stock. Your account architecture should connect channels commercially without pretending they do the same job.
A clean hybrid structure separates activity by two questions: how aware is the customer, and where should they complete the purchase?
At the top of the funnel, use creator content, paid social video and broad audience activity to introduce the product, problem or category. These campaigns may send traffic to your own site when education, lead capture, bundles or subscription economics make that worthwhile. For some brands, the best destination will be Amazon because marketplace trust and Prime delivery produce a higher conversion rate. There is no universal answer. Test the destination rather than following a channel preference.
In the middle of the funnel, focus on consideration. This is where product benefits, comparison messaging, proof points and review-led creative matter. Retarget video viewers, site visitors and social engagers with a more direct product proposition. Keep Amazon and DTC retargeting audiences distinct where possible. A customer who has viewed a product page on your site is not identical to an Amazon shopper who has searched the category.
At the bottom of the funnel, protect demand aggressively. Brand search, Amazon branded terms, product retargeting and competitor defence should be structured as separate, tightly monitored activity. Do not allow generic prospecting campaigns to consume budget intended to defend high-converting branded demand.
The destination rule needs to be explicit. If Amazon pricing is lower than your website, sending high-intent shoppers to DTC may suppress conversion. If your site offers higher-margin bundles, better first-party data and strong post-purchase retention, DTC may be the more valuable destination even with a lower initial conversion rate. Measure contribution margin, not vanity revenue.
Within Amazon, structure campaigns around controllable levers. Mixing discovery, defence and harvesting in one campaign makes optimisation slower and hides where money is being wasted.
A practical architecture normally separates automatic campaigns for search-term discovery, manual keyword campaigns by match type and intent, product targeting campaigns, branded defence and competitor conquesting. Sponsored Brands and Sponsored Display should sit alongside this structure with a defined role, not as a collection of experimental campaigns left to run unchecked.
For example, a high-volume hero ASIN may require separate campaigns for branded exact terms, generic category exact terms, phrase and broad exploration, competitor ASINs, complementary product ASINs and retargeting audiences. That level of separation lets you set bids according to expected conversion and margin. A branded search term that converts at 20% should not be governed by the same bid logic as a broad category term converting at 6%.
Do not over-segment simply to make an account look sophisticated. If a product has limited conversion data, splitting it into dozens of campaigns creates noise, not control. Campaign granularity should increase with spend, search volume and strategic importance. The objective is decision-quality data.
Stock status belongs in the advertising structure too. There is little value in scaling traffic to an ASIN with weak inventory cover, an unstable Buy Box or poor retail readiness. Advertising cannot compensate for a detail page that lacks reviews, has weak imagery or sits at an uncompetitive price.
Hybrid accounts need a measurement hierarchy. Platform ROAS and ACoS are useful operational metrics, but neither should be the final budget decision-maker.
At Amazon level, monitor ad sales, ACoS, TACoS, conversion rate, new-to-brand signals where relevant, organic rank and total sales movement. At DTC level, track blended MER, contribution margin after media and fulfilment, customer acquisition cost, repeat purchase behaviour and email or SMS capture. Across the business, assess total revenue and total paid spend by product line, not simply by platform.
The key question is incremental profit: what sales would not have happened without this spend? You will not answer that perfectly from a dashboard. Use controlled tests where practical. Reduce spend on a selected audience, region, product group or campaign type, then compare the movement in total sales rather than only reported channel sales. Watch for Amazon sales rising when Meta spend increases, or DTC branded search increasing after Amazon awareness activity. Those relationships are often commercially real even when attribution platforms fail to connect them.
Create one weekly scorecard with channel spend, attributed revenue, total Amazon revenue, DTC revenue, blended media efficiency, contribution margin and stock cover. This prevents a common failure mode: cutting an upper-funnel channel because its reported ROAS is weak, then discovering six weeks later that branded search and Amazon conversion have deteriorated.
A hybrid budget should not be fixed by channel for an entire quarter. Set a base allocation for proven activity, then reserve a meaningful test budget for new audiences, creative, products and destinations. The scale of that reserve depends on maturity, but brands that allocate nothing to testing eventually plateau.
Pacing rules turn strategy into action. Define thresholds before spend rises: the maximum acceptable ACoS by product margin, the minimum contribution margin for DTC acquisition, the stock cover required before scaling, and the conditions under which a campaign is paused or expanded. A campaign should not be scaled because it had two good days. Equally, it should not be cut after a short period if it supports a wider demand-generation objective.
Review budgets by product portfolio, not only by channel. Your hero products may justify aggressive investment because they create repeat buyers or lift the wider range. Low-margin products may need defensive coverage only. Seasonal lines require a different pacing model again, with spend building ahead of the peak rather than reacting after competitors have taken the demand.
The first is duplicated targeting with no ownership. Meta retargeting, Google remarketing, Amazon DSP or Sponsored Display and email can all chase the same small audience. Without frequency controls and a clear sequence, you pay several times to reach a customer who was already ready to buy.
The second is forcing every channel to meet the same ROAS target. Discovery activity, branded defence and high-intent conversion are fundamentally different investments. Apply different targets, but make each target commercially defensible.
The third is optimising advertising without retail context. Falling conversion can be caused by price, stock, reviews, delivery promise or content quality. A senior operator looks beyond bid changes before deciding that traffic is the problem.
Finally, avoid reporting that celebrates isolated platform wins. A lower Amazon ACoS is not a victory if total Amazon revenue is falling. A high Meta ROAS is not a victory if it is merely claiming customers driven by existing brand demand.
The strongest hybrid accounts are not the most complicated. They give each channel a defined role, protect profitable demand, test the right destination and judge success against total contribution. When the numbers stop competing for credit, you can direct spend towards the products and audiences that genuinely grow the business.