Why the percentage rule is the wrong start
“Spend 5 to 10% of revenue” is repeated everywhere and it answers the wrong question. It tells a £200,000 trade business to spend £10,000 to £20,000 a year without saying on what, or whether one more job would cover it. For a business whose next customer is worth £3,000 in margin, a channel costing £200 an enquiry is a bargain at any percentage; for one whose customer is worth £30, the same channel is ruinous at any percentage.
People ask how much marketing agencies charge, and the honest answer is “what the client will bear”, which is why the question should be turned round: what can the business afford to pay for a customer, given what a customer is worth?
Start from one number: what a customer is worth
- Average job or order value, in profit not revenue. A £2,000 driveway with £600 margin is a £600 customer, before repeat work.
- How many times they buy, and how many they refer. A sweet shop customer may reorder monthly; a driveway customer may refer two neighbours. Estimate conservatively.
- Multiply: that is roughly what a customer is worth over their life with you.
- Decide what share of that you would pay to acquire them. A third is common; a business with capacity to fill can go higher, one at capacity should go lower.
Where the first pounds go
| Order | Spend on | Roughly | Why first |
|---|---|---|---|
| 1 | Google Business Profile and a review habit | Free; an afternoon and a template | The map pack sits above everything on local searches and keeps working |
| 2 | A website that converts: price, phone, one action, speed | From £500 build, £50 a month care (our prices) | Every later channel multiplies the conversion rate |
| 3 | Measurement: Search Console, conversion tracking, an enquiry log with sources | Free tools; a few hours to set up | Without it you fund the channel that feels busiest |
| 4 | Pages that rank for buying searches and answer buyer questions | Your evenings, or a growth block from £250 a month | Compounds; the cost per enquiry falls every month |
| 5 | Adverts, once 2 and 3 are in place | Whatever the sum from the one-number exercise allows | Instant volume you control; stops when you stop |
Source: Cybrial’s working order with clients; prices correct at 5 September 2026.
What help costs, and how it should be charged
Agencies charge in four ways: a monthly retainer (defined or not), a project fee, a percentage of ad spend, or a per-lead fee. Only two are sound. A project fee for a defined piece of work, and a retainer that names its hours and its deliverables. A percentage of ad spend pays the agency for spending more, and can run to 40% of the budget. A per-lead fee without a definition of a lead pays for form fills.
Our model is the second kind: a block of hours from £250 a month, spent on whichever of the five rows above moves the needle, with a plain summary of what the hours went on and what the numbers did, and no minimum term. For a business at the smaller end, that is the whole marketing budget beyond ad spend, and it is deliberately sized so that one extra job a quarter covers it.
A worked example
A trade business with a £600 average margin, one referral in three, so roughly £800 per customer. Willing to pay a third, £270, to acquire one. Currently ten enquiries a month converting one in three: three jobs, £1,800 margin. The budget that makes sense is whatever produces additional enquiries at under £270 each with the current conversion rate, or improves the conversion rate so that existing enquiries produce more jobs.
In that business, £250 a month on the profile, the site and the pages is one extra job a quarter to break even, and everything beyond that is return. £1,000 a month on adverts to a site converting one visitor in a hundred is not; the same £1,000 after the conversion work is. The order matters more than the total.
