On this page
Find the shape of your own year
Before reading anyone's opinion about when your trade is busy, plot your own. Export the last twenty-four months of invoices, total them by calendar month, and put them in a chart. It takes twenty minutes and it is more useful than any industry article, because it describes your customers in your area rather than a national average that includes London.
Two cautions on that chart. Twenty-four months is a small sample, so one unusually large contract can make a month look like a peak when it was a coincidence — check whether the shape repeats in both years before believing it. And invoice dates lag the enquiries that generated them, often by weeks, which is the single most important adjustment and the subject of its own section below.
What you are looking for is not precision. You want to identify your two or three genuine peaks, your reliable trough, and the four to six weeks before each peak when the buying decisions are actually being made.
Do the same exercise a second time with the revenue split by job type, because the aggregate curve often hides two opposite ones. A firm doing both roofing repairs and full re-roofs will find the repair line spikes with bad weather in the winter while the re-roof line peaks in late spring, and the total flattens the pair of them into something that looks like a business with no season at all. Those two halves need different budgets, different ad copy and different pages, and the aggregate chart will never tell you that. The same is true wherever domestic and commercial work sit in one business: they usually run on different clocks, and public sector work runs on a third.
Where the data comes from
- Your invoices, two years by month. The ground truth for revenue, though it lags demand.
- Your enquiry log, if you have one, which is closer to the real demand curve than revenue is. If you do not have one, starting it is the highest-value thing in this guide.
- Search Console, which holds sixteen months of impression data. Impressions for your service terms are a decent proxy for local demand and cost nothing.
- Google Ads impression data by month, if you have been running long enough, which shows the auction rather than just your share of it.
- Google Trends, five years, filtered to the UK and to your region where the sample allows. Remember it reports relative interest on a 0–100 scale rather than absolute volume, and at regional level for a niche term the line is noisy enough to mislead. Use it for the shape, never for sizing.
Where these disagree, trust your own enquiry log first and Search Console second. National trend data is the weakest of the sources and the one most often quoted as though it were the strongest.
The lag that catches everyone out
Marketing has to lead revenue by the length of your pipeline, and most firms discover this by advertising into a peak that has already been booked out elsewhere.
Work it out explicitly. Take the average time from enquiry to quote, add the average time from quote to acceptance, add the lead time from acceptance to work starting. If you quote within four days, customers take a fortnight to decide and you start three weeks later, your total lag is around six weeks. That means the marketing that fills your April has to be running in the middle of February, and the budget increase has to happen before that, because bidding strategies take a couple of weeks to settle after a change.
Organic work has a far longer lag again. A new service page realistically takes three to six months to reach a position where it earns meaningful traffic. The corollary is uncomfortable but useful: the pages that will win you work next April need writing in October or November, which is exactly when everyone feels least like doing it. That is the real argument for keeping organic work running through the quiet months rather than pausing it.
Weather-led against calendar-led trades
The two behave completely differently and need different plans.
Weather-led trades — roofing, drainage, gutters, fencing, tree work, groundworks — get step changes in demand within hours of a named storm or a hard frost, decaying over roughly a week. You cannot forecast the date, so the plan is structural: keep a small emergency campaign running all year rather than building one when the wind picks up, have the ad copy written and approved in advance, and be ready to multiply a daily budget for four or five days. Firms that build the campaign after the storm arrive for the expensive tail, bidding against everyone else who had the same idea.
Calendar-led trades have demand you can diary. Heating and boiler work climbs from September as the first cold snap reminds people. Garden and landscaping enquiries build from late February through May. Extensions and larger domestic projects are decided in January and February for summer builds. Commercial and public sector work bunches ahead of the March and April financial year end, when unspent budget needs committing.
Most firms are a mixture, which is an argument for separating the two into distinct campaigns with distinct budgets rather than one account that half-serves both.
A twelve-month plan
| Period | What tends to happen | What to be doing about it |
|---|---|---|
| January | Domestic enquiries for planned work return in the second week; emergency work spikes with frost | Publish the pages that need to rank by spring. Set the year's budget shape. |
| February | Spring decisions are being made now | Raise budget on planned work. This is the lead time for April. |
| March | Commercial and public sector year end; domestic picks up | Push commercial campaigns hard. Chase old quotes. |
| April–May | Peak for most external trades | Full budget, watch impression share lost to budget weekly. |
| June–July | Busy, often capacity-constrained | Trim secondary services rather than cutting everything. Collect reviews while jobs are completing. |
| August | Holidays flatten decision-making | Maintain, do not panic. Good month for photography and page work. |
| September | Heating and interior work turns on | Shift budget between services. Refresh ad copy. |
| October | Last window for external work; storms begin | Emergency campaigns ready. Write the pages for next spring. |
| November | Interior and urgent work only | Reduce planned campaigns, keep emergency live. |
| December | Two dead weeks, then domestic emergencies over the break | Set special hours everywhere. Small emergency budget over the holiday. |
Adjust the whole table by your own lag and your own curve. It is a starting frame, not a prescription, and a firm doing commercial refurbishment has a different year from one doing domestic bathrooms.
Moving money rather than adding it
The instinct in a trough is to cut spending, and in a peak to raise it. Both are half right and both are usually mistimed.
The better model is a fixed annual marketing budget with a shifting mix. Through the peak, the money sits in harvesting demand: search campaigns, tight geography, maximum budget on the campaigns that convert. Through the trough, it moves into building demand and building assets — the content and technical work that ranks by spring, photography from the jobs you have just finished, review collection, and the paid social that works better for considered, higher-value projects than for emergencies.
Two mechanical points. Move budget two to four weeks before you want the effect, because automated bidding re-paces after a change. And do not cut a campaign to zero if you intend to bring it back, because you lose the recent conversion history the bidding depends on and pay for a re-learning period. Reducing to a floor is almost always better than pausing.
There is also a capacity question hiding inside the budget question. Raising spend into a peak you cannot service produces quotes you send late, jobs you book too far out, and customers who go elsewhere while holding a poor opinion of you. If your constraint in June is crews rather than enquiries, the money is better spent lifting the value of the work you take — bidding harder on the higher-value services, raising prices, and letting the cheaper enquiries go — than on generating more of the same. Marketing that outruns delivery costs you reputation as well as margin.
What the quiet months are for
- Write the pages that need six months to rank. The single highest-return quiet-month activity.
- Photograph and document finished work while you can still remember what was involved. Everything on your site next year comes from this.
- Chase the quotes that went cold. A quiet week is exactly when the follow-up you never do becomes affordable.
- Fix the site problems you have been ignoring since April. Speed, forms, the pages nobody has read since they were written.
- Talk to the trades that feed you work. Relationships with builders, merchants and specifiers do not have a season and are usually the first thing dropped when you are busy.
- Look at what next year needs against what the sector pages and your own numbers suggest is realistic.
The traps
Judging a channel on its worst month. Every channel looks terrible in the trough. Compare against the same month last year or you will cancel something that works.
Spreading the annual budget evenly. A twelfth every month is administratively convenient and commercially wrong, because it underinvests when enquiries are cheap to convert and overinvests when nobody is buying.
Reacting to the peak instead of leading it. By the time you notice the phone ringing, the enquiries for that period are already placed.
Turning everything off in December. The two weeks around Christmas are genuinely dead for commercial work and unusually good for domestic emergencies, at low competition. A small emergency budget over the break often pays better than the same money in March.
Treating one hot summer as a pattern. Two years of data is thin. Where your own numbers and the trend line disagree, wait another season before rebuilding the plan around it.
If you want help building the curve for your own business, send over twelve months of enquiry data and we will tell you what shape it is and where the money should move. Doing that first is part of how we work, and it usually changes the answer on budget more than any change to the campaigns themselves.
Rather have someone else do this?We run exactly this for trade and construction firms across the UK. Call and we will tell you what is realistic in your area, including when the answer is that it is not worth it.
Call 07443 392243WhatsAppCommon questions
Should I stop advertising completely in my quiet season?
Reducing is usually better than stopping. Pausing loses the recent conversion data that automated bidding relies on and adds a re-learning period when you restart, which lands exactly as your season begins. Drop to a floor budget on your best campaign and move the rest into work that pays off later.
How far ahead of my busy season should I increase budget?
By your own lag, which is the time from enquiry to quote, plus quote to acceptance, plus acceptance to starting work, plus a fortnight for bidding to settle after a change. For many trades that totals six to eight weeks, meaning the April increase happens in February.
Is Google Trends reliable for a local trade?
It is useful for the shape of a year and unreliable for anything else. It reports relative interest rather than search volume, and at town level for a specific trade term the sample is often too small to be stable. Use your own enquiry log and Search Console impressions as the primary sources.
What if my trade has no obvious season?
Some do not, particularly compliance-driven work and reactive commercial maintenance. Plot the data anyway, because you may find the season is in something other than volume, such as a shift between domestic and commercial mix, or a change in average job value across the year.
Does seasonality affect SEO the same way it affects ads?
The demand curve is the same but the response time is completely different. Advertising responds in days, so it can chase a season. Organic work takes months, so it has to anticipate one. That is why the content for your busiest quarter should be published two quarters earlier.