Google Ads agency fees, explained
Four pricing models. Real UK numbers. And the one reason percentage-of-spend is more expensive than it looks — even when the sticker price is lower than a fixed fee.
Nobody wants to write this article. Agencies do not, because it exposes how our own market is priced. Advertisers do not, because they do not have the data. So the guides that rank for "how much does a Google Ads agency cost" are either advertising or vague.
Here is the actual answer, from someone who runs a specialist retainer model and sees competitor pricing every week.
The four pricing models
Percentage of spend
AvoidTypical: 8-15% of monthly ad spend
Feels 'aligned' at first glance.
Misaligns incentives on day one. The agency's revenue only grows if your spend grows, regardless of whether that spend is profitable. Recommends 'scale' when the correct recommendation is 'pause'.
Fixed monthly retainer
RecommendedTypical: £2k-£15k/month depending on scope
Agency has no financial incentive to inflate spend. Decisions are commercial, not budget-preserving. Predictable P&L line.
Needs a minimum spend floor to be viable for the agency (typically £10k-£15k/month ad spend). Below that, the maths do not work for anyone.
Performance / commission
Rarely worth itTypical: Retainer + % of revenue or profit above baseline
Fully aligned on paper. Correct incentive if the profit definition is real (CM3, not revenue).
Almost never structured correctly. 'Performance' usually means revenue or blended ROAS, both of which the agency can game without moving profit. Also creates perverse incentives around attribution.
Hourly / project
SituationalTypical: £100-£250/hour, £5k-£30k projects
Works for one-off audits, migrations, or consultancy engagements. Clean scope, no ongoing commitment.
Wrong shape for retained media management. PPC is a compounding discipline, not a project. Hourly billing incentivises complexity, not simplicity.
Why percentage-of-spend costs more than the number on the invoice
Take a £50k/month advertiser on a 10% arrangement. Sticker price: £5k/month, £60k/year. Same advertiser on a fixed £5k/month specialist retainer pays £60k/year. Identical, right?
Not once you look at incentives. The percentage model earns 10p more for every £1 of ad spend, regardless of whether that pound made or lost money. So over 12 months, the recommendations skew towards "let us scale into this new market", "let us add another PMax campaign", "let us test another bid strategy at higher spend". Some of those are right. Some are not. You will not be able to tell which.
The fixed retainer earns £5k whether spend is £30k or £80k. So the recommendations come out differently: "these SKUs should be capped", "PMax should be shrunk to two asset groups", "pause this campaign for 60 days". Same operator, same skill level, entirely different bias.
Over 12 months on a £50k/month account, that bias is worth £30k-£80k in avoided waste. Which makes the percentage model, at identical sticker price, materially more expensive than fixed fee.
The commercial test
Ask any prospective agency this: "If, six months in, the right recommendation is to pause 30% of the account and reduce total spend by 40% for two quarters — will your fee structure reward you for saying that?"
Only one pricing model can answer yes without hesitation. That is the model you want.
What good pricing actually looks like in 2026
- 01Fixed monthly retainer, priced against scope not spend
- 0230-day rolling notice after an initial 3-month term
- 03Written scope covering strategy, execution, reporting and access
- 04Client-owned accounts, tags, feed and data. No lock-in
- 05Named specialist on the account, not an AM triaging to an operator
- 06No fee-tier ladder based on spend brackets — the work does not double when spend doubles
Frequently asked questions
What is the average Google Ads agency fee in the UK?
For SME ecommerce (£15k-£100k/month ad spend), typical fixed retainers sit at £2k-£8k/month. Percentage-of-spend arrangements typically land at 10-12%, which at £50k/month spend is £5k-£6k — usually more expensive than a comparable fixed fee, with worse alignment.
Why is percentage of spend still the most common model?
It is easy to price, easy to justify at pitch stage ('you pay more when you spend more, so we are aligned'), and it scales the agency's revenue automatically as clients grow. It persists because it benefits agencies, not because it benefits advertisers. The alignment argument does not survive contact with a P&L.
What is the minimum sensible retainer?
For fixed-fee retained management, roughly £2k/month. Below that, either the client is underserved (a couple of hours a week, junior operator, no strategy) or the agency is running at a loss and will churn the account. Anyone quoting £500-£1,500/month for retained Google Ads is misrepresenting the scope.
How much should I be paying at £50k/month ad spend?
Fixed retainer: £4k-£8k/month for a specialist. Percentage-of-spend equivalent: £5k-£7.5k. Full-service agency: £6k-£12k. The specialist fixed retainer is usually the best commercial deal and the best strategic fit for that spend level.
Are cheap agencies actually cheap?
No. A £1.5k/month retainer that mismanages a £50k/month account costs you the retainer plus the roughly £150k/year in leaks that a proper operator would find. The retainer is a rounding error against the spend it manages. Optimise for who manages the spend, not who charges least to do it.
Should I ever pay for pitches?
Yes, when the pitch includes real analysis of your account. A paid audit (£500-£2,500) filters out agencies who cannot do the diagnostic work and gives you a document you own regardless of whether you sign. Free pitches select for salespeople, not operators.
What about performance-based fees tied to profit?
The idea is right; the execution is almost never right. To work, the contract must define profit at CM3 level, agree a baseline from a specific historical period, and lock the attribution window. Most 'performance' deals define profit as revenue or blended ROAS, both of which move without profit moving. If the agency will not commit to CM3, the model is not what it says on the tin.
Do I need a long contract?
No. A 30-day rolling notice after an initial 3-month term is the industry standard for good specialists. Anything longer (12-month lock-ins with penalty clauses) is a red flag: the agency is protecting itself against churn instead of protecting the relationship with performance.
What should the retainer actually cover?
Strategy (bid architecture, SKU roles, break-even ROAS), execution (bid management, feed, PMax structure, exclusions), reporting (CM3 after paid, mix drift, incrementality), and access to a specialist (not an account manager forwarding questions to an operator). If it covers less than that, the price is not the issue — the scope is.
Continue:
- • How to choose a PPC agency — the scorecard
- • Should I change PPC agency? — the six commercial questions
- • Grade your current account — 15 questions, 3 minutes
Fixed fee. £2k/month floor. No % of spend.
Book a profit audit to see what a specialist retainer would find in your account.
Book a profit audit