There is still a common assumption among growing brands that the right answer to Amazon is to hire someone capable, give them the channel to own, and trust that the rest will follow. It is an understandable instinct. Headcount feels controllable, accountability is clear, and the cost looks contained on paper. For brands that did this five or six years ago, it was usually the correct decision. The platform was simpler, the talent pool deeper, and the work largely repeatable once someone had learnt the basics.
That window has effectively closed.
What has changed is not really the principle of in-house ownership, which remains valuable in plenty of areas of a business. What has changed is the shape of Amazon itself, the maturity of the talent market around it, and the speed at which decisions now must be made to stay competitive. Each of those shifts is significant individually. Together, they make the single-hire model almost impossible to execute well, even with budget behind it.
The Hiring Reality Most Brands Underestimate
Recruiting for Amazon roles has become disproportionately difficult relative to the seniority of the position. To give some scale to that, our last open role attracted 313 applications. Twelve of those progressed to a phone screen, four reached a structured interview, and even within that final group the skill alignment we needed was not really there. The shortlist was made up of competent ecommerce people, but very few had the depth across advertising, catalogue, operations, and reporting that a modern Amazon account manager actually requires.
The implication for a brand hiring directly is uncomfortable. If a specialist agency reviewing 313 applications struggles to find a strong shortlist, an internal hiring team without daily exposure to the platform is unlikely to fare better. The result tends to be a hire who looks reasonable on paper but who is two or three years behind where the platform now is. Training closes some of that gap, but only with significant investment, and only when there is someone internally qualified to actually train them. Most brands do not have that person, which is why training tends to be informal at best and accidental at worst.
The cost of getting this wrong is not just salary. It is twelve to eighteen months of slowed momentum on what is probably the largest share of the brand’s online revenue.
Amazon Has Become More Complex Than Every Other Marketplace Combined
The other shift, which is harder to appreciate from outside the channel, is just how much the platform has changed in the last five years. Advertising alone now contains more campaign types, placement logic, audience options, and reporting surfaces than most full ecommerce platforms have in their entirety. Catalogue rules, brand registry mechanics, A+ content variants, attribution windows, DSP, and the operational side of FBA each carry their own learning curves and their own ways of going wrong.
It is genuinely fair to say that Amazon today is more complex than every other marketplace combined. eBay, B&Q, Shopify, even direct-to-consumer infrastructure, all sit at a level of operational depth that an experienced ecommerce person can pick up reasonably quickly. Amazon does not. It rewards continuous exposure and punishes occasional attention, which is exactly why a single hire stretched across the whole surface area tends to optimise the parts they are most comfortable with and quietly under-serve the rest.
What makes this awkward is that Amazon also remains too important to handle imperfectly. It accounts for around 43% of UK ecommerce, and for a meaningful share of brands, it is the first place a customer encounters them. The first impression of a brand on Amazon often ends up being the impression of the brand overall, which means weak execution on the channel is rarely contained to that channel alone.
The Question That Tends to Expose Outdated Thinking
A question we ask in almost every account management interview is straightforward. If you had a thousand active ASINs, all with stock, none with strong seasonality, how would you decide what to work on first.
The answers are remarkably consistent, and that is the problem. Candidates talk about looking at bestsellers, sorting by Amazon’s listing quality score, or running everything through a Helium 10 listing score and starting at the bottom. None of those are wrong exactly. They were reasonable starting points in 2022. They are no longer adequate as a primary prioritisation method in 2026, and the fact that they remain the default answer says something about how the industry has trained people to think.
The deeper issue is that all three approaches begin from a static snapshot. They look at where a listing currently sits rather than at the trajectory it is on, the contribution it is producing, the competitive dynamic around it, or the cost of inaction. A bestseller might already be at peak efficiency with no further upside. A listing with a low quality score might be a low-volume SKU where the work would never pay back. The signal that actually matters is rarely the one any single index surfaces.
Data Has to Lead, or Speed Disappears
For a brand with significant Amazon revenue, the bottleneck is increasingly not effort, but the speed at which decisions get made. There is more to look at than there used to be. More dashboards, more reports, more places where something might be quietly going wrong. The practical consequence is that without data doing the prioritisation work, the team ends up reacting to whichever issue happens to be most visible that week, rather than the one that matters most commercially.
Modern Amazon account management has to be built around data that tells you, by SKU, where contribution is moving, where competitive position is shifting, where stock cover is becoming risky, where conversion has slipped, and where the largest opportunity exists relative to the effort required to capture it. That is a different shape of work to “go through the catalogue and see what looks weak”, and it is not a shape that one stretched hire can produce on top of everything else they are responsible for.
The brands moving fastest at the moment are not necessarily the ones with more headcount on Amazon. They are the ones whose decision-making cycles are shorter, because the data layer underneath has been built to surface the right priorities automatically.
Ads Are Ahead, Account Management Is Catching Up
One observation worth making is that the advertising side of Amazon is materially further along this curve than the account management side. Bid management, placement optimisation, search term harvesting, and budget pacing have been moving into automation and rules-based logic for some time. The discipline now expects data to lead, and most experienced PPC operators would find it strange to work any other way.
Account management is still mostly manual, mostly judgement-led, and in many businesses still organised around the same rhythms it had years ago. That gap will close, and the brands that close it first inside their own operation will move ahead of those that do not. The frustrating part is that closing it requires either an internal data and operations capability that very few brands have, or a partner whose model is built around it from the outset.
Either way, the in-house single-hire approach sits awkwardly in between. The person carrying the role is unlikely to have the time, the tooling, or the supporting infrastructure to operate at that level, regardless of how capable they are individually. That is not a reflection on the hire. It is a reflection on a structure that no longer fits the platform it is meant to manage.
For most brands at scale, the realistic question is no longer whether to keep Amazon in-house or move it elsewhere. It is whether the model around the work, whichever side of the line it sits on, is actually designed for how the channel operates in 2026.
