TL;DR
- A price tells you what leaves your account. It tells you nothing about what comes back.
- The only question worth asking a supplier is how long the work takes to pay for itself, and what has to be true for that to happen.
- Payback period is the one calculation a small business can actually do. The fancier ones mostly add decimal places to a guess.
- Some of the return genuinely cannot be attributed. Say so, rather than inventing a number for it.
Three quotes land on your desk. One is half the price of the others. It is very hard not to feel that the cheap one is the sensible, prudent choice, and very easy to forget that you have just compared the only number that says nothing about whether the work is worth doing.
Price is an input, not a verdict
A price is a fact about your bank account on a particular Tuesday. It has no opinion about whether the thing you bought will earn anything. Two quotes of the same size can behave completely differently once the invoice is paid: one becomes an asset that keeps producing enquiries for three years, the other becomes a maintenance obligation you are still paying for in year two.
This is not an argument for spending more. Plenty of expensive work is bad work, and agencies have an obvious interest in telling you that cheap is a false economy. It is an argument for asking a different question — not “what does it cost?” but “what has to happen for this to pay for itself, and how would I know?” If nobody can tell you what the work has to produce to break even, the price is the only thing being sold.
The one calculation worth doing
Payback period is the amount of time it takes for a piece of work to return what it cost. It is crude, it ignores the time value of money, and it is the only one of the standard measures a small business will realistically keep up. Do this one properly before worrying about the others.
A worked payback calculation
- Start with the true cost
Not the quoted fee. Add hosting, licences, the retainer, and the days of your own time the project will eat. A quote of eight thousand pounds with a monthly retainer of four hundred is a first-year cost of nearly thirteen thousand.
- Find the value of one customer
Average order or contract value, times how many times a typical customer buys, minus what it costs you to deliver. If a client is worth two thousand pounds of margin over two years, that is your unit.
- Work out how many you need
Thirteen thousand divided by two thousand is between six and seven new clients to break even in year one. Write that number down. It is the whole business case in one figure.
- Sanity-check it against reality
Six or seven extra clients from a new website is plausible for a firm currently winning forty a year. For a firm winning eight, it is a claim that the work will nearly double the business, and somebody should have to defend that out loud.
- Agree how you will know
Decide before the project starts what you will count — enquiry forms, tracked calls, booked meetings — and who is checking. A return nobody measures is a return nobody can claim.
Most of the value in that exercise is in step four. It converts an abstract argument about quality into a specific, testable claim that either survives contact with your own sales figures or does not.
Where the return actually shows up
Two parts of a web investment have genuinely documented returns, and it is worth knowing which they are, because they are rarely the parts that get discussed in the pitch.
The first is speed. In Milliseconds Make Millions, research commissioned by Google and carried out by 55 and Deloitte across 37 brands in 2019, a 0.1 second improvement in mobile site speed moved 3.2% more shoppers from the product listing to the product page and 9.1% more from the product page to the basket. Retail spend rose 9.2% and travel bookings 10%. That is retail and travel data and should not be read straight across to a professional services site — but it establishes that performance work has a measurable financial return, which most design decisions cannot claim.
The second is the checkout, if you sell online. The Baymard Institute, averaging across 50 studies, puts the documented online shopping cart abandonment rate at 70.22%. That is an ecommerce figure and means nothing for a brochure site, but where it applies it identifies exactly where the money is going: not the homepage everyone argues about, but the three screens at the end that nobody wants to pay to fix.
Notice what both have in common. They are specific, they are measurable, and they are unglamorous. That is generally what a real return looks like.
Where the maths stops working, and why you should say so
This is the part the usual version of this article leaves out. Net present value and internal rate of return are genuinely useful in capital budgeting, where the cash flows are contractual. Applied to a marketing budget, they take a number you guessed — how many enquiries a new site will bring — and process it into a figure with two decimal places. The precision is entirely manufactured. It has the effect of making an assumption look like a finding.
Worse, a real proportion of marketing spend cannot be honestly attributed at all. Somebody sees your work, remembers the name eighteen months later, searches for you directly and phones. Your analytics will record that as a direct visit and give the credit to nothing. Any model that claims to have accounted for it is guessing, and any supplier who insists otherwise is selling certainty rather than measurement.
So read any business case put in front of you against the same four questions, and notice which version of each answer you are being given.
| What to look at | An honest business case | One to distrust |
|---|---|---|
| The headline number | One clear break-even figure you can check against your own sales | An uplift expressed as a percentage, with no source |
| The assumptions | Stated as assumptions, in plain words | Projections carried to two decimal places |
| The evidence | A named metric, and a named person checking it | “Industry benchmarks” that name no industry |
| Attribution | An admission of what cannot be traced | Credit claimed for everything that happens next |
Deciding under uncertainty is normal, and it is not an excuse for the kind of arithmetic that pretends the uncertainty is gone. Better to say: this needs to bring in six clients, here is why we think it can, and here is how we will know by March.
What to ask before you sign
Take these to whichever supplier you are talking to, ourselves included. The answers matter less than whether the questions are welcome.
A supplier who has thought about the return will answer these in plain language and will usually volunteer the awkward parts before you ask. One who has not will move the conversation back to the price, which is where the cheapest quote always wants the conversation to be.
Good to know
FAQs
No. Sometimes the cheap option is genuinely the right fit, particularly if you need to be online quickly and the stakes are low. The mistake is comparing prices without first working out what the work has to earn.
Work out the full first-year cost, work out what one new customer is worth to you in margin, and divide. The result is how many customers the work has to bring in to break even. If that number looks unreasonable against your current sales, the project needs rethinking.
Rarely. Those methods assume you know the future cash flows. On marketing you are estimating them, and running an estimate through a discount rate makes a guess look like a finding without making it any more accurate.
State the gap rather than filling it. Track what you can genuinely track, ask new enquiries how they found you, and treat the untraceable share as a known limit on your certainty rather than a number to invent.
In the unglamorous parts: how quickly pages become usable, and how few obstacles sit between an interested visitor and getting in touch or completing a purchase. Those are also the parts most often left out of a redesign quote.
Next step
Comparing quotes and not sure what you are comparing?
We work with businesses across Brighton and London and will tell you what a project has to earn before you commit to it.