Full disclosure before a word of comparison: rankhouse is one of the three options this piece compares, so read it knowing where we sit. We've tried to earn your trust the only way that works in writing, by being specific about the weaknesses of our own model and the genuine strengths of the alternatives, and by telling you plainly who shouldn't hire us. We've worked agency-side, worked alongside agencies, and worked freelance next to MDs and FDs, so the failure modes below are all ones we've seen from the inside.
Option one: the large agency
The genuine strengths. Capacity and coverage. A large agency can staff a 400-ASIN catalogue across five marketplaces tomorrow, has process for everything, survives any individual leaving, and brings pattern recognition from a big client roster. For enterprise brands that need industrial-scale execution and procurement-friendly structure, these are real advantages, not brochure copy.
The characteristic failure mode. The delivery pyramid. The person who wins your business is rarely the person in your campaigns by month three; that's a junior running forty accounts to a playbook, reviewed monthly by someone senior. Nothing scandalous, it's how the economics of scale work, but it produces a predictable texture: competent maintenance, slow reactions, reporting built to demonstrate activity, and a strategy that resembles every other client's. The incentive question to ask any percentage-of-spend model is the one we'd ask: if the agency earns more when you spend more, who in the building is paid to want your ads efficient?
| Large agency | Boutique specialist | In-house hire | |
|---|---|---|---|
| Who runs the account day to day | Often a junior on many accounts | Senior specialists on a small client base | Your hire, your training curve |
| Cost shape | Retainer, often % of ad spend | Fixed monthly fee | Salary, tools and management time |
| Per-ASIN attention | Spread thin by design | Deliberately limited client numbers | Full, once they know the platform |
| Incentive alignment | % of spend rewards spending more | Fixed fee rewards results, not spend | Aligned, but a single point of failure |
| Best suited to | Very large catalogues needing volume | Brands wanting senior attention on profit | Brands with budget for a proven operator |
Option two: in-house
The genuine strengths. Total alignment and accumulated context. A good in-house operator lives your margins, sits in your forecasting meetings, and compounds brand knowledge no external partner fully matches. At sufficient scale, typically when the channel justifies a small team rather than one hire, in-house is the correct end-state for many brands, and we'll say so even when it costs us renewals.
The characteristic failure mode. Concentration risk and ceiling risk. One hire means one perspective, no cover for holidays or departures, and skills that age exactly as fast as Amazon changes, which is fast: this year alone brought a restructured fee stack, an AI search transition and a Prime Day that jumped its slot. The salary maths also surprises founders: a genuinely senior operator costs more than most retainers, and a junior one costs less while learning on your account. The honest hybrid many brands land on: in-house ownership of the channel, external specialist depth on the parts that move.
Want this level of thinking inside your account?
rankhouse is a boutique specialist agency: experienced Amazon professionals, no juniors, led personally by founder Dan Whalley. Fixed fee from £1,500 + VAT a month. Every engagement starts with a free profit audit built from your own data.
Get the free audit at rankhouse.co.uk →Option three: the boutique consultancy
The genuine strengths. Senior attention and aligned incentives. In a founder-led boutique, the person you evaluated is the person in the account, daily, and a fixed-fee model means efficiency is the product rather than a threat to revenue. Decisions happen at the speed of one conversation. On accounts where per-product economics, not headcount, are the constraint, this model tends to win on results per pound, and it's why we built rankhouse this way: small client base, founder in the account, fixed fee, never a percentage of spend.
The characteristic failure mode, stated against ourselves. Capacity is the hard limit. A boutique that takes every client stops being one; the model only works because it says no. That means a good boutique may not have room when you call, can't staff a 400-ASIN, five-marketplace catalogue on industrial timelines, and concentrates key-person risk of its own. And the market makes it worse: "boutique" is an easy word to print, so the burden falls on you to check whether senior attention is the operating model or the sales page. Ask who, by name, touches the account weekly, and how many clients that person carries.
The comparison that actually decides it
- Incentives: percentage-of-spend pays for growth in spend; fixed fees pay for results per pound; salaries pay for tenure. Each shapes behaviour exactly as you'd predict.
- Resolution: whoever runs the account, demand reporting at the altitude where truth lives: profit per unit, TACoS per product against its own breakeven, never blended. Any partner who can't produce that resolution is managing your revenue, not your business, and the Qualkem story is what the difference looks like in money.
- The named human: the single best predictor of outcomes in this industry is who, specifically, thinks about your account and how often. Get the name, the cadence and the client count in writing.
Who shouldn't choose us, in plain terms
We keep a page on our site titled who we're not for, and it applies here. If you need enterprise-scale headcount tomorrow, a large agency serves you better. If you want a partner who'll promise a number in the first call, plenty will; we won't, and our targets model explains what we do instead. And if the channel has outgrown external help entirely, we'll tell you, and help you hire. The right answer to this comparison changes as brands grow. The wrong answer is choosing by pitch deck instead of by incentives, resolution and the named human, and that test, usefully, can be applied to everyone. Including us: the free audit at rankhouse exists so you can judge the resolution of our thinking on your own data before a penny changes hands.
Questions we get asked about this
What should a percentage-of-spend agency fee make me ask?
One question, asked politely and insistently: who in your building is paid to want my advertising smaller? Percentage models tie the agency's revenue to your spend, which doesn't make agencies villains, but it does mean efficiency, the kill lists, the bid discipline, the honest 'spend less on this product', works against their income. Watch for the tells: reporting that celebrates spend deployment and ROAS rather than profit per unit, resistance to per-product breakeven analysis, and growth plans that always require bigger budgets. Some percentage agencies do excellent work despite the incentive; the point is the incentive never helps you, and a fixed fee removes the conflict rather than asking people to be noble about it.
At what point does in-house become the right answer?
When the channel can justify a small team rather than a single hire, because the single hire carries the model's two structural risks alone: concentration, one perspective, no cover, everything walking out the door in one resignation, and ceiling, skills ageing exactly as fast as Amazon changes, which this year alone meant a restructured fee stack, an AI search transition and a Prime Day that jumped its slot. The salary maths matters too: genuinely senior operators cost more than most retainers, juniors cost less while learning on your account. The pragmatic path many brands run: external specialist depth building the systems and the playbook, an in-house owner grown alongside it, and a planned handover when scale justifies the team. We'll help brands make that transition; it's the honest end-state for some of them.
What questions actually expose whether a 'boutique' is real?
Four, all with checkable answers. Who, by name, works in my account weekly, and is it the person I'm evaluating now? How many clients does that person carry, because senior attention divided by forty accounts is a pyramid wearing a boutique's clothing? Can I see the reporting resolution on a real anonymised account, per-product profit, TACoS against breakeven, or is it a blended dashboard? And what happens when you're at capacity, because a boutique without a waiting list answer is one that takes everyone, which is the same as no model at all. The pattern in the answers matters more than any single one: specificity is the tell of the real thing.
How do I compare costs fairly across the three models?
Price the resolution and the attention, not the invoice, because the invoice is the smallest number in the comparison. A cheap retainer that misses a mispriced hero product, an unclaimed fee backlog or a quietly ad-dependent catalogue costs multiples of its saving; a senior operator, in any model, who finds those things pays for years of fees in one finding. The fair comparison: total cost of the option, fee or salary plus tools plus management time, against the resolution of decision-making it buys, per-product economics or blended averages, and the speed of reaction it buys, daily attention or monthly review. Then sanity-check against your account's leak profile, because the more unaudited the account, the more the senior option is worth.
Can I mix the models rather than choosing one?
Most scaled brands eventually do, and the good mixes are designed rather than accumulated: external specialist depth on the moving parts, fee structures, AI transitions, event strategy, expansion, with in-house ownership of brand context and cross-channel coordination; or a boutique running the account with a large agency's production arm handling volume creative; or in-house operators with an external senior review cadence as insurance against the ceiling risk. What makes mixes work is the same trio that decides the single-model choice: clear incentives for each party, one shared reporting resolution so everyone argues from the same per-product truth, and named humans with explicit ownership. What makes them fail is overlap nobody mapped, three parties, one blended dashboard, and no one accountable for profit.
The next step is twenty minutes.
If any of this reads like your account, the fastest way to find out is the free audit: per-product profitability, where the ad spend is leaking, and what we would fix first. No pitch deck, no obligation.
Book a chat with Dan →
About the author
Daniel Whalley is the founder of rankhouse, a boutique specialist agency for Amazon-focused growth in FMCG, health, wellness and beauty brands. He has spent 10 years inside Amazon accounts, generating £100M+ for the brands he works with, and manages £500k+ a month in ad spend across the UK, Europe and the US. He writes from inside the accounts he runs, not from the sidelines. Connect on LinkedIn → · amazon@rankhouse.co.uk