How to Plan for Your Quiet Season Using Last Year's Traffic

Every business has a stretch of the year where things go thin. For a bakery it might be the weeks after Christmas. For a solicitor, August, when half the country is away and nobody is moving house. For a plumber, the mild middle of summer, when nothing has frozen or failed. For a freelance designer, whenever clients' budgets reset and sign-off goes quiet.

The problem is rarely that the quiet season exists. It is that most of us notice it from the till, the diary or the bank balance, which all lag behind reality by three to six weeks. By the time the slow month feels slow, the decisions that would have softened it — the email you should have sent, the page you should have written, the shift you should not have rostered — are already behind you. Your website usually knows first. People stop looking before they stop buying, and that gap is where the planning happens.

Open your analytics and set the date range to the last thirteen to twenty-four months. Thirteen is the minimum, because it lets you compare this January to last January rather than to December, which tells you nothing useful. Then look at one number only: unique visitors per month. That is the count of separate people who came to your site, rather than the count of pages they looked at, so one person reading four pages counts once. If you want the detail of how that is worked out, the unique visitors documentation covers it in plain terms.

Write the monthly figures down somewhere you can see them all at once. A sheet of paper is fine. Suppose a bakery's year reads something like 2,100 in November, 3,400 in December, 1,900 in January, 1,450 in February, 1,700 in March. The eye goes straight to December, because big numbers are satisfying. Ignore December. The number that matters is 1,450, and the thing you are planning for is the slope that leads into it.

Do this for two years if you have them. One quiet February is weather. Two quiet Februaries in a row is a season, and a season is something you can plan against with a straight face.

Now switch the same chart from monthly to weekly. Monthly figures smooth the edges off everything and hide the moment the slide actually begins. Weekly figures are noisier but far more honest about timing.

What you are hunting for is the last normal week — the final week before the line starts stepping down and does not recover. In the bakery example, the bottom is mid-February, but the weekly view might show the decline starting in the second week of January. That five-week gap is the entire point of the exercise. It means next year, the useful date in your calendar is not "February is quiet" but "by the first week of January, the quiet-season plan is already running".

Do the same for the recovery. If traffic reliably picks up again in the third week of March, you know when to stop spending defensively and start spending to catch the wave. Both dates matter. Marketing into a dip that still has four weeks to run is an expensive way to feel busy.

A drop in traffic is not always a season. Before you commit money to it, rule out three dull explanations. First, did your tracking break? If a site redesign or a plugin update knocked the measurement code off some pages, you will see a sharp cliff rather than a gentle slope, and often on specific pages rather than across the board. Seasons slope; breakages cliff.

Second, did one channel fall while the others held steady? A genuine quiet season usually shows up everywhere at once — search, social, direct visits, email. If search traffic halved while everything else stayed flat, that is a search problem wearing a season's coat, and it needs a different response entirely.

Third, was there something you did? A newsletter that went out in November and not in January will move the numbers all by itself. This is why it is worth marking events on your traffic chart as they happen — a price change, a local radio mention, a site migration, a holiday closure. Annotations let you pin a note to a date so that next year's you is not staring at a mysterious bump trying to remember what happened. It takes ten seconds and saves an hour of guessing.

Different marketing has wildly different lead times, and the mistake almost everyone makes is reaching for the slowest one at the latest possible moment. Writing a new page to rank in search is a reasonable thing to do; expecting it to bring in visitors within a fortnight is not. Search work typically takes a couple of months to show anything, sometimes longer.

So order your actions by how fast they work and count backwards from the week the slide starts. Email to your existing list is the fastest lever most small businesses have — people already know you, and you can send on Tuesday and see bookings on Wednesday. Paid ads are fast but cost money, and a quiet season is when you can least afford a lazy campaign. An offer or a packaged service sits in between: it needs a week or two to write, photograph and publish.

A plumber whose quiet stretch begins in early June might land on a plan like this: mid-April, write and publish a boiler service page; mid-May, email the customer list offering a pre-summer service slot at a sensible price; first week of June, a small, tightly targeted ad campaign for the same thing. Nothing clever. Just everything happening four to six weeks earlier than instinct would have allowed.

Traffic shape is a staffing input, with caveats. It tells you about interest, not about revenue, and the two are not the same — a solicitor's August visitors might be fewer but more serious. Use the shape as a prompt to look at your own booking or sales records for the same weeks, and let the two together inform the rota. If both say the third week of February is the floor, that is where the training day, the deep clean or the annual leave goes.

The quiet season is also the only time of year you will realistically do the website jobs that keep getting pushed back. Broken links, slow-loading pages, a contact form nobody has tested since the last update, pages that exist but nothing links to. These are not urgent in November and they are never urgent, which is how they survive for years. If you want a list of what is actually wrong rather than a vague sense of unease, the free report card scans your site and tells you, in order, which is a more useful starting point than opening your homepage and squinting at it.

One honest limit: analytics cannot tell you that a competitor opened down the road, that your best supplier put prices up, or that the council dug up the street outside. The traffic shape shows you what happened, not always why. Keep your own notes alongside it.

Here is the awkward timing problem. The content that earns you visitors in February needs to be published in November or December, when you have the least time to write it. There is no way round this except to accept it and schedule it.

Work out what people search for in your quiet months, which is often different from your busy months. The bakery's December searches are about Christmas orders; the February ones are more likely to be about birthday cakes, gluten-free options, or simply "bakery near me" from someone who has run out of bread. If you have connected your site to Google's search data, you can see the actual phrases people used to find you last February rather than guessing — the Search Console integration brings those terms in alongside your visitor numbers.

Then write two or three pages in your busy season, aimed squarely at the quiet one, and publish them with enough runway that search engines have found and settled on them before you need them.

All of this collapses into about ten lines: the week the slide starts, the week it bottoms out, the week it recovers, and three or four dated actions with a name against each. Put it somewhere you will see it — pinned above the desk is better than buried in a folder.

The failure mode is not the plan. It is October arriving and nobody remembering the plan exists. So attach it to something that turns up on its own. A weekly summary that lands in your inbox is a reasonable nudge; the weekly report flags changes in your numbers and suggests what to do about them, ranked, which at least means someone is looking at the trend every Monday even when you are elbow-deep in the week. Set a calendar reminder four weeks before your slide date with the plan attached to it, and that is genuinely most of the job done.

Next year, do the whole exercise again with two years of data instead of one. It gets quicker and considerably more accurate each time.

Plenty of people read this, open their analytics and find three months of history. That is not a failure, it just means this year you plan on softer evidence and next year you plan on real evidence.

In the meantime, use what you do have. Your bank statements show the shape of your year even if your website does not. So does your booking diary, your invoice dates, your supplier orders. Ask two or three people who have been in your trade locally for longer than you have — they will tell you when the quiet weeks are, and they will usually be right. Then start recording properly now so that this time next year the guessing stops.

The main thing is to get measurement running before the season you want to understand, not after it. You can start on the free plan and simply let it collect while you get on with the actual business. A year of honest data costs you nothing but the decision to begin, and it is the only way to stop treating a predictable dip as a nasty surprise.