I run an agency, so read this knowing that. I’ve tried to write the version I’d want to read if I were on the other side of the table, which means starting with the thing that undercuts my own pitch.
There is no published study comparing Amazon PPC agency performance against in-house performance. I looked. Every result is an agency or a software vendor arguing its own corner, with no methodology and no sample. If someone shows you a chart proving agencies deliver X% better ROAS, ask who paid for it.
That absence is worth sitting with, because it means the decision can’t be made on outcome data. It has to be made on cost structure and on what each model tends to get wrong.
What each one actually costs
Agency, UK market:
| Tier | Monthly retainer |
|---|---|
| Freelancer or VA | £200–£500 |
| Independent UK agency | £1,250–£3,500 |
| Full-service mid-market | £3,500–£8,000 |
| Premium full-service | £8,000–£16,750+ |
Alternative models: 10–20% of ad spend, or 3–10% of monthly revenue. Multi-marketplace expansion typically adds 30–50% for the US and roughly doubles for the EU. Onboarding or audit fees run £2,000–£10,000 as a one-off.
The 10–20% of ad spend band is the one figure that shows up independently across UK and US sources, which makes it the safest anchor. It’s also consistent with the 8–15% separately reported for DSP management.
In-house, UK:
| Role | Salary | Day rate |
|---|---|---|
| Amazon PPC Executive | £35,000–£45,000 | £200–£300 |
| Amazon PPC Manager | £38,000–£55,000 | £300–£400 |
| Senior Amazon PPC Manager | £55,000–£75,000 | £375–£500 |
| Head of Amazon PPC | £70,000–£85,000 | £500–£750 |
Those are placement figures from a specialist UK ecommerce recruiter, which makes them better sourced than most numbers in this debate.
Now gross them up. Employer National Insurance plus a minimum 3% pension adds roughly 15–18% to gross salary. A £48,000 PPC Manager costs about £56,000–£57,000 before tooling, recruitment fees, holiday cover or the management time of whoever they report to.
The tooling line people forget
Helium 10 Platinum is $129/month, Diamond is $359/month, or $99 and $279 billed annually. Jungle Scout runs $29–$129/month depending on tier.
Then read this, from Helium 10’s own pricing page: “Diamond plan customers incur a 2% management fee on PPC spend managed through Helium 10 Ads.”
That’s the line that catches brands out. On £40,000 of monthly ad spend, that 2% is £800 a month — more than double the subscription itself. Your tooling bill scales with your account, not with your plan.
It’s the same structure as percent-of-spend agency pricing, and it deserves the same scrutiny. Perpetua starts at $695/month for up to $10,000 of spend with an undisclosed percentage above that, on annual contracts requiring 90 days’ written notice before renewal. Pacvue publishes no pricing at all; third parties estimate roughly $500/month minimum or about 3% of spend, whichever is higher.
A realistic UK in-house stack is £200–£800 a month, plus any percent-of-spend fees layered on top.
The honest cost comparison
A brand spending £30,000 a month on Amazon ads:
| In-house | Agency (15% of spend) | |
|---|---|---|
| Salary, loaded | £4,700/mo | — |
| Tooling | £400/mo | usually included |
| Helium 10 Ads 2% fee | £600/mo | — |
| Recruitment, amortised over 2 years | £250/mo | — |
| Management time | not costed, but real | — |
| Retainer | — | £4,500/mo |
| Total | ~£5,950/mo | £4,500/mo |
At £30,000 of spend, an agency is cheaper on paper. Push spend to £80,000 and the percentage model flips: 15% is £12,000 a month against an in-house cost that barely moves.
That crossover is the whole economic argument, and it’s about scale rather than quality. Below roughly £40,000 monthly spend, an agency usually costs less than a competent hire. Above it, in-house wins on cost — provided you can hire and keep the right person, which is where the argument stops being about money.
Where each model actually fails
Since there’s no performance data, the useful question is what goes wrong.
The agency failure mode is incentive misalignment. Percent-of-spend billing pays the agency more when spend rises, whether or not margin does. That’s not an accusation of bad faith; it’s just what the contract rewards. Ask any agency quoting a percentage what happens to their fee when the right answer is to cut spend by 30%.
There’s a second one, and the data on it is genuinely interesting. A 2026 benchmark survey of 494 agency professionals found the top reasons clients leave: budget cuts 42%, client-side internal changes 37%, perceived value erosion 32%, performance misses 31%.
Read those last two together. Clients leave agencies over value perception more often than over results. What that tells you as a buyer is that an agency can be doing good work and still lose you, usually through poor communication, which means the reporting cadence and the quality of explanation you get is not a soft factor. It’s the thing most likely to determine whether the relationship survives.
The same survey found 68% of agencies run operating margins above 20%, with 60–79% of team time billable. Useful for backing into how much of your retainer is actually delivery.
The in-house failure mode is single-point coverage. One person owns the account. They go on holiday, they get ill, they leave. Amazon PPC is not a discipline you hand to a colleague for two weeks.
And there’s a subtler version that’s become sharper this year. Look at what Amazon shipped in 2026 with no advertiser action required:
- Off-Amazon creator placements, 10 August 2026: existing campaigns enrolled automatically at current bids
- Prompt Ad Extensions — Amazon’s own words, “zero additional setup required”
- Auto-assembled videos attached to existing Sponsored Products ads
- DSP automatic inventory expansion, where opting out requires contacting a consultant rather than a self-serve toggle
- A new Agent Policy in the Business Solutions Agreement, effective 4 March 2026, governing automated tools accessing your account
Somebody has to read Amazon’s release notes. If nobody in your business does, you find out about these changes from a variance in your spend report, weeks later. A single in-house manager with a full workload rarely has that time. A good agency has it as a shared cost across clients, which is the strongest structural argument for the model, and it’s about attention rather than talent.
What actually decides it
Four questions, and none of them is “which is better.”
1. What’s your monthly ad spend? Under £40,000, an agency is usually cheaper than a loaded hire. Above it, run the arithmetic on your own numbers.
2. Can you hire and keep the person? A £38,000 executive won’t have the judgement for a complex account. A £55,000 manager will, and will also be recruitable by everyone else. If you can’t retain them, the hire is a rolling recruitment cost with a knowledge gap in the middle.
3. Who reads the release notes? Whoever owns the account needs several hours a month on Amazon’s changes alone — the 2026 fee restructure and the title rule both landed with no warning to most brands. If that’s not funded, it doesn’t happen.
4. What’s the fee model? If it’s percent-of-spend, ask what happens when the correct answer is less spend. If the answer is unconvincing, the incentives are wrong regardless of how good the team is.
The hybrid that works well in practice: an in-house owner who holds the commercial context, the margin data and the product roadmap, plus an agency for architecture, platform change monitoring and the specialist work that doesn’t justify a second salary. It costs more than either alone. It also fails less often than either alone.
What I’d tell you if you weren’t a prospect
If your Amazon business is under about £500,000 a year, hiring a dedicated PPC manager is hard to justify and an agency retainer at the lower end will do more for you.
If you’re over £3 million and Amazon is your primary channel, you should have someone in-house who owns it, whether or not you also use an agency. At that size the account is too central to your business to sit entirely outside it.
Between those two, it depends on the four questions above rather than on anything an agency tells you.
FAQ
How much does an Amazon PPC agency cost in the UK? Independent agencies typically charge £1,250–£3,500 a month, mid-market full-service £3,500–£8,000, and premium £8,000+. Percent-of-spend models cluster at 10–20%.
What does an Amazon PPC manager earn in the UK? £38,000–£55,000 for a manager, £35,000–£45,000 for an executive, £55,000–£75,000 senior. Add roughly 15–18% for employer National Insurance and pension to get the true cost.
Is an agency cheaper than hiring in-house? Below roughly £40,000 of monthly ad spend, usually yes, once salary is grossed up and tooling is counted. Above that, percent-of-spend models scale faster than a fixed salary and in-house tends to win on cost.
What tools does an in-house Amazon PPC manager need? Around £200–£800 a month for a UK stack. Watch for percent-of-spend fees layered on subscriptions — Helium 10 charges 2% of PPC spend managed through Helium 10 Ads on top of the Diamond plan.
Is there evidence agencies outperform in-house teams? No published study with a disclosed methodology exists. Every comparison in circulation is produced by a party with an interest in the answer, including this one.
Sami Sultan runs Ecom Enable, an Amazon growth partner for DTC brands, so this article has an obvious interest in its own conclusion. Salary, tooling and agency pricing figures verified 27 August 2026 and attributed to their sources so you can check them.
Sources: 3Search 2026 UK ecommerce salary guide; Helium 10 pricing; AgencyAnalytics 2026 Marketing Agency Benchmarks (n=494); Mr. Prime UK Amazon agency cost survey; Amazon Ads API release notes; Amazon Business Solutions Agreement update, March 2026.