The quarter that decides the year

The last quarter of the year is the one that decides how the year went, for shops and for anybody earning from advertising. The reasons are mechanical and the preparation is almost entirely a timing problem, which is why most people get it wrong by starting in December.

Two separate things happen at once

They get discussed as one and they are not, and separating them is what makes the preparation obvious.

• People buy more. Gift buying concentrates a year of discretionary spending into a few weeks, and more of it happens online than at any other point in the year.
• Advertising gets more expensive. Every brand wants to reach those buyers, so what advertisers bid rises. For a shop that is a cost. For a site carrying advertising it is income.

So a shop faces higher demand and higher acquisition costs simultaneously, while a publisher simply earns more per visitor. If you do both, the two move against each other and should be budgeted separately.

Why advertising income rises, precisely

Google's own documentation explains the mechanism: what you earn per click is market driven and depends on what advertisers are bidding, and total revenue is a product of that price, the click rate and the number of impressions. Nothing about your site changes in the fourth quarter. The price advertisers are willing to pay does.

The pattern of a fourth-quarter peak and a first-quarter trough is reported consistently in the published results of advertising companies, which describe budgets peaking in the fourth quarter and the first quarter being the lowest. So the January drop is not something you did wrong, and Google's guidance for diagnosing earnings changes is explicitly to compare year on year before assuming a fault, because a pattern that repeats is usually outside your control.

The timing problem

Content published in December does not rank in December. A new page needs to be found, assessed and given a position, and on a site without much authority that takes weeks at best. Meanwhile buyers start researching well before they buy, so the searches you want to appear for are happening before the spending does.

1. 01Publish seasonal content by early autumn. Not because October is magic, but because it gives the page time to be indexed and to collect its first signals before demand arrives.
2. 02Update last year's pages rather than writing new ones. An existing page that already has a position is worth more than a new page with none, and updating it is a fraction of the work.
3. 03Check stock and supplier lead times before promoting anything. A ranking page for a product you cannot ship is worse than no page.
4. 04Make sure the site is quick on a phone before the traffic arrives, because the quarter is where the cost of being slow is highest.
5. 05Write the January content in December. Gift cards, returns, what to do with what you were given: that demand is real and arrives when nobody has prepared for it.

What not to do when rates are high

• Do not add advertising units because each one is worth more. A page that becomes unpleasant loses the reader and the next visit, and Google's publisher policies treat ads that interfere with content as a violation in their own right.
• Do not discount reflexively. Demand is already there, and a discount in the one quarter where people are willing to pay full price is margin given away for nothing.
• Do not promise delivery dates you cannot hold. In this quarter a missed date is not an inconvenience, it is a present that did not arrive.
• Do not change your advertising layout or your site structure in November. You will not be able to attribute anything that happens next.

The part people skip

Write down what happened while you still remember it. Which pages brought the traffic, which products sold, what ran out, what arrived late, which supplier answered the phone. In January that costs an hour and next September it is the entire plan. Almost nobody does it, which is why most people rebuild the same preparation from memory every year and make the same two mistakes.