How to Create a Digital Marketing Budget
A percentage of revenue is a guess dressed as a policy. A figure built from your target, your margin and your own arithmetic can be defended in a meeting, adjusted when things change and reviewed against what it produced. This is how to arrive at the second one.
The Two Ways Budgets Get Set
Almost every marketing budget is arrived at by one of two routes. One is quick and defensible only by convention. The other takes an afternoon and survives being questioned.
Percentage of revenue. A share of turnover, set annually.
Simple to calculate, easy to approve and it makes budgeting a single decision rather than a project. It also has one genuine merit: it scales with the business automatically.
Objective and task. Start from what you need to achieve.
Define the commercial goal, work out what has to happen to reach it and cost that. Slower, though it produces a figure connected to something real.
Which is better. The second, clearly.
A percentage tells you nothing about whether the amount is enough, too much or pointed at the right things. Two businesses at the same turnover can need very different budgets, which block three demonstrates.
Why the easier one persists. It needs no evidence.
Objective and task requires knowing your close rate, your conversion rate and your margin. Businesses that do not have those figures reach for the percentage. The percentage then hides the fact that they are missing.
Working Backwards From The Target
A chain of five links, each derived from the last. Every input is a fact about your own business rather than an industry figure, which is what makes the answer defensible.
The chain. Target, sale value, close rate, enquiry rate, cost per visit.
A worked example. An illustration of the method, not a benchmark.
Every figure below is invented to show the arithmetic. Replace all of them with your own and ignore these entirely.
A business wants £150,000 of additional revenue. Assumption: the average sale is £3,000, so it needs 50 sales. Assumption: one enquiry in five becomes a sale, so it needs 250 enquiries. Assumption: two visitors in a hundred enquire, so it needs 12,500 visits. Assumption: those visits cost £2 each to acquire, so the budget required is £25,000.
What the chain gives you. A conversation with the target.
If £25,000 is unaffordable, the arithmetic shows exactly which lever to pull. Improve the enquiry rate and the traffic requirement falls. Improve the close rate and the enquiry requirement falls. Or reduce the target.
Where the inputs come from. Your records, not an article.
Anything you cannot evidence gets written into the plan marked as an assumption, so that when the result differs everybody can see which guess was wrong.
Cost Per Lead And What You Can Afford To Pay
The previous block tells you what a target costs. This one tells you whether you can afford it. It is also the section most budget discussions never reach.
Allowable cost per acquisition. The ceiling.
The most a customer can cost to win before the sale stops being worth making. It is set by your gross margin rather than by your revenue, which is the distinction everything else here rests on.
Why revenue misleads. A second worked example.
Both figures below are invented for illustration. Two businesses each turn over £500,000. The first works at a forty percent gross margin, so a £3,000 sale contributes £1,200. The second works at ten percent, so the same sale contributes £300.
What follows. The same customer is worth four times as much to one of them.
The first can spend meaningfully to win a customer and still profit. The second cannot spend a quarter of that. Identical turnover, entirely different affordable budgets.
What repeat purchase does. Raises the ceiling.
If a customer typically returns, the allowable cost is set against everything they will spend rather than the first transaction. That is a legitimate and dangerous adjustment, since it commits money now against revenue that has not happened.
Why copying a competitor fails. Their margin is invisible.
Their spend tells you nothing without it. A channel comfortably profitable for them can be unaffordable for you.
Fixed And Variable Costs
Most small marketing budgets count media spend, sometimes a retainer, then discover the rest during the year as unbudgeted requests.
Retainers and fees. Fixed, monthly, predictable. The easiest to remember and usually the only fixed cost anybody writes down.
Software and tools. Email platform, scheduling, analytics, call tracking. Individually small, collectively significant, accumulating quietly.
Media spend. The variable half. What actually gets paid to platforms for reach.
Production of assets. The one that gets forgotten. Writing, design, editing, building pages.
Photography and video. Frequently the single largest one off cost. Frequently unbudgeted entirely.
Internal time. Yours and your staff's, at whatever an hour of it is worth.
Why the omission matters. Production is what gets advertised.
A budget that funds promotion with nothing to promote ends up amplifying whatever already existed, however out of date. The media spend then performs poorly. The media gets blamed for it.
Splitting The Budget
Split it three ways: capturing demand that exists, creating demand that does not, plus a fixed slice held back for testing. We publish no percentages for those, because the right split depends on things that differ between businesses.
Demand capture. Being found by people already looking.
Converts fastest and is capped by how many people are actually searching. Beyond that ceiling, more money buys nothing.
Demand creation. Reaching people before they look.
Slower, harder to attribute and the thing that fills next year's pipeline rather than this month's.
The test slice. Held back, always.
Money set aside for things you do not yet know will work. Without it a budget calcifies into whatever worked two years ago.
What shifts the balance towards capture. Thin margins, short sales cycles, existing awareness, urgent need for cash.
What shifts it towards creation. Healthy margins, long consideration periods, a category people do not know exists, else a demand ceiling you have already hit.
Why we will not give you a rule. Any published split is somebody's average.
Applied to your business it is a coincidence rather than an answer. It will be wrong in a direction you cannot see.
Building Or Buying
Four routes, each with a real advantage and a real cost. We are an agency, so treat this section with appropriate suspicion and check that we have given each option its due.
In house. Someone employed to do it.
They learn your business properly and are available continuously. The costs are salary, tools, cover during absence, management time and the fact that one person cannot be expert in more than two or three disciplines.
Freelance. Specific skills, bought by the day.
Often the best value per hour, with genuine expertise in one thing. The costs are availability, continuity when they take other work and the management burden falling on you.
Agency. A team, bought as a service.
Several disciplines at once, cover during absence and accountability in one place. The costs are real: a margin on everything, less familiarity with your business than an employee would have, the risk of being a small account among larger ones, plus the fact that a retainer keeps being charged whether or not the work that month was worth it.
A mix. What most small businesses end up with.
Someone internal owning it, with specialists bought where the skill is genuinely specialised. This is usually the cheapest route. It only works if the internal owner has actual time.
Cash Flow And Timing
Marketing spend and marketing results happen at different times. Budgeting as though they coincide produces a predictable and avoidable problem.
The lag. Spend now, enquiries later, cash later still.
Even fast channels have a gap between the money leaving and the invoice being paid. For anything with a long sales cycle, that gap can span a quarter.
What businesses do in a quiet month. Cut spending.
Understandable and it is the decision that extends the quiet period. Reducing spend in a slow month produces another slow month later, at which point the cut looks justified rather than causative.
Seasonality. Spend before the season, not during it.
Demand arrives in waves and the work has to be done before the wave. A trade spending on advertising during its busiest month is paying to reach people it cannot serve.
How to plan around it. Budget annually, spend unevenly.
An annual figure divided into twelve equal amounts ignores everything above. Weight the spending towards the weeks before demand arrives, then hold a reserve for the quiet period so the cut never has to be made.
Testing Money And Stop Rules
Every budget should contain money for things that might not work. What separates a test from a slow loss is a rule written before it starts.
What a test slice is for. Buying information.
Not results. The output of a test is knowing whether something works, which is worth paying for and should be budgeted as a cost rather than expected to return.
Define success in advance. Specifically.
What has to happen, by when, for this to continue. A test with no defined success is a channel that gets kept or killed on how somebody feels at the review.
Write the stop rule first. The most important line here.
The point at which you stop, decided while nobody is emotionally committed. Written afterwards, it becomes a negotiation with money already spent.
Why afterwards fails. Sunk cost.
Once real money has gone in, stopping means accepting the loss and continuing feels like recovering it. Almost everybody continues, which is how a test becomes a subscription.
Reviewing And Reallocating
A budget set annually and never revisited is a forecast rather than a management tool. Reallocating well means knowing when the evidence is sufficient.
The cadence. Quarterly, for money.
Frequent enough to correct a mistake within the year, infrequent enough that no channel gets judged on noise.
What justifies moving money. A difference in cost per enquiry, sustained.
One month of difference is variation. A quarter of consistent difference is evidence. The size of the gap matters as much as its direction.
What does not justify it. A quiet month, a competitor's launch, an article somebody read.
The discipline that gets broken. Leaving a channel alone.
Every channel needs an uninterrupted run to produce a readable result. Moving money monthly means nothing ever completes a fair run, so nothing is ever proven either way and the budget churns without improving.
What a fair run is. Different per channel.
Weeks for paid advertising, quarters for search and content. Agreed in advance, as set out in the marketing plan.
When The Budget Is Very Small
With little money, order matters more than at any other budget level, because there is no room to run several things at once and see what happens.
First, the website and the tracking. Costs little, changes everything.
A site that converts and enquiries that get recorded. Every pound spent afterwards works harder. You can also tell which pounds worked.
Second, claim the free profiles. Time rather than money.
The business profile, the directories that matter in your trade, complete and current. For local businesses this is frequently the highest return available anywhere.
Third, capture the demand already searching. Cheapest customers there are.
People looking for what you sell. The wanting already happened, so you are paying only to be found.
Fourth, email the customers you have. Nearly free.
People who already bought are the least expensive sale available. Almost every small business neglects them entirely.
Fifth, buy reach. Once something is proven to convert.
Advertising multiplies what exists. With a small budget, multiplying something unproven is how the whole budget disappears with nothing learned. What to watch is in return on marketing investment.
Common Mistakes
Each of these is a budget that looks complete on the page while missing something structural.
Media with no production money. Funding promotion without funding what gets promoted. The advertising then carries whatever already existed. The media takes the blame.
No test allocation. Every pound committed to what already runs. The budget then repeats last year indefinitely, which feels safe while quietly becoming out of date.
Buying reach before the site converts. Paying to send more people to something that loses them. The commonest expensive mistake in this list, also the most avoidable via an audit.
Matching a competitor's spend. Their margin sets what they can afford. Yours sets what you can. The two are invisible to each other. Display buying in particular varies enormously in what it can justify, per what is display advertising.
Treating the figure as fixed for twelve months. A budget that cannot move cannot respond to evidence. The whole series is on the digital marketing guide.
