Why your Q4 budget should not be one number

Every guide published this month will tell you not to set a single Q4 budget. Model scenarios instead. Protect October. Hold a reserve. Write your scaling rules down before the quarter starts.

All of it is correct, and we would give you most of it ourselves.

None of it explains why businesses keep setting one number anyway, which is the part worth understanding, because in Australia the reason is structural rather than lazy.

Why this matters

The Australian timing problem nobody writes about

Almost all Q4 advice is written for the northern hemisphere, where the calendar year and the financial year end together. Set the annual budget, and the peak quarter is the one immediately in front of you.

Here the financial year ends on 30 June. Which means the biggest trading quarter of the year sits in the middle of your financial year, and the budget funding it was approved in May or June, roughly five months before anyone could see what November was going to look like.

That is where the single number comes from. Not carelessness, but a planning cycle that asks for a figure long before the information exists.

It also explains why the conversation in October so often becomes a request for more budget rather than a plan. The number was set in a different context, by people who are not watching the account, against a forecast made before the year's trading was known.

The fix is not to argue for a bigger number. It is to convert the number you already have into a shape, and to agree the conditions under which it moves.

The number that makes one figure fail

Here is the mechanism, and it is the most useful idea in this whole subject.

Average return on ad spend hides what the next dollar does. A campaign generating $600,000 from $100,000 in spend is running at a 600% return. Push it to $125,000 and generate another $100,000, and the account now shows $700,000 from $125,000, still a healthy 560% average. But the incremental $25,000 returned 400% (Vizion Interactive).

The average stays reassuring while the marginal return falls away underneath it.

This is why a single budget figure is dangerous in a peak quarter. It commits you to a spend level before you can see where your own marginal return gives out, and in Q4 that point arrives sooner than usual because the auction is more expensive.

Which is also why the useful pull-back trigger is contribution margin per order rather than ROAS. When margin per order sits below your floor at the current spend level for more than a couple of days, you have moved past the efficient frontier for that audience (Webtopia).

What to give them instead of a number

Three things, and they can all be prepared now.

A shape. Take last year's revenue week by week and align the calendar properly, so Black Friday lands on Black Friday (Foxwell Digital). Two businesses can put the same money into the quarter and end up somewhere completely different depending on when it went out (Razor Rank).

A range, not a point. Model conservative, expected and aggressive side by side with the order volumes each implies, because the business owns the risk tolerance, not the media buyer, and a choice between three costed options is a decision someone can actually make (Foxwell Digital).

A written rule for moving. What evidence justifies an increase, how big the increment is, who makes the call, and what triggers a pull back. Decisions made against written rules at nine in the morning are consistently better than decisions made against a dashboard at eleven at night (Webtopia).

What people overestimate

Last year's curve. It is the right starting point and it is not clean data. A week that looks like a demand spike might have been a closeout, a one-off bundle or pricing you have since retired, so interrogate the shape before you build on it (Razor Rank). Treat old numbers as the shape of the seasonal curve and the levels as fiction (Adscalr).

The published cost increases. Search for what Q4 costs and the figures contradict each other by several times over, because each source averaged a different pile of strangers, and none of them measured your account (Adscalr). Sizing the rise is the wrong first job.

Rising costs as bad news. A CPC moving from $2.00 to $2.75 looks like a problem until the conversion rate moves from 3% to 4.5%, at which point the media got dearer and the quarter got better (Vizion Interactive). Only one of those two facts usually makes it into the meeting.

More budget as the lever. Before asking for more, check whether campaigns already converting at an acceptable cost are losing impression share to budget, because moving money from weaker activity into proven campaigns often produces more volume than increasing the total (Adzooma). Lifting a landing page from 4% to 5% is 25% more conversions from the same traffic, which is usually a better trade than buying more clicks at November prices (Adzooma).

Ecommerce advice applied to lead generation. Almost everything written about Q4 budgets assumes you are selling products. If you sell services, the ceiling is not your budget, it is how many enquiries your team can actually follow up (Vizion Interactive). Buying more leads than you can call is not growth, and December is genuinely quiet for most B2B, so the shape of a lead gen Q4 looks nothing like a retailer's.

What to do in the next fortnight

Work out the shape before you argue about the size.

Pull last year's revenue by week, align the calendar, and mark where the money actually came in rather than where you assumed it did. Build three columns rather than one, and put the order volume and the margin next to each so the business can pick with its eyes open. Then write down, while nothing is on fire, what evidence lets you increase spend, by how much, and what makes you stop.

Hold something back. Not a token contingency, a real reserve, because the whole point is to fund the days that surprise you.

And if your financial year ended in June, accept that the number you were given was a guess made in autumn about a summer nobody had seen yet. That is not a reason to ignore it. It is a reason to go back now, with a curve and three scenarios, rather than in the third week of November with a request.

At Dadek Digital we build this plan with businesses in September, so the quarter runs against a shape and a set of rules rather than a single figure and a lot of nerve. If you would like us to stress test your Q4 spend plan against your actual margins before the quarter starts, we are happy to take a look.