How Does Seasonality Affect SEO Performance Data?
A report that ignores seasonality will produce a wrong conclusion twice a year in most businesses. Once when the quiet period arrives and looks like failure. Once when the busy period arrives and takes credit for work nobody did.
Nearly Every Business Has A Shape
Seasonality is usually thought of as something affecting shops at Christmas. In practice it applies to almost every trade, frequently more sharply than it does to retail. Most businesses have never written their own pattern down.
Why it goes unrecognised. The word.
Seasonal sounds like weather and holidays. A business whose demand is driven by contract renewals or house moves does not think of itself as seasonal at all, while having a shape just as pronounced.
How sharp it can be. Sharper than retail.
A shop's busy period is a multiple of its quiet one. Some trades effectively have work and no work, which makes the same measurement mistake far more damaging.
What it does to a report. Two wrong conclusions a year.
The quiet period reads as the work failing. The busy period reads as the work succeeding. Neither reading has anything to do with what was done.
Why it belongs in this series. It is a measurement principle.
Not a marketing observation. Until you know your shape, no period of data can be read correctly and no comparison means anything.
The test. Name your quiet month.
If you cannot, that is a gap in the record rather than evidence you have no pattern. Block seven covers establishing it.
The Shapes Are Not All Weather
Demand cycles come from several different sources and they behave quite differently from each other. Assuming weather is the only one is why so many businesses fail to recognise their own pattern.
Weather driven. The familiar kind.
Heating in the cold months, drainage after heavy rain, outdoor work in dry spells. Gradual, roughly predictable and shifted by an unusual year.
Fixed calendar dates. Sharper.
Anything tied to a deadline that falls on the same date annually. These produce a spike rather than a curve. The spike is reliable to the week.
Academic and tenancy years. Frequently missed.
Anything touching students, rented property or institutions moves on an academic rhythm rather than a calendar one. A whole industry can be quiet in the middle of summer.
Property transactions. A knock-on cycle.
Trades whose work follows people moving house inherit the property market's shape rather than having one of their own, which makes their pattern a copy of somebody else's.
Renewal dates. The quietest one.
Anything on an annual contract produces demand when those contracts expire. Invisible unless you look for it, then precisely repeating once you do.
Why the distinction matters. Predictability differs.
A weather cycle shifts by weeks between years. A date driven one does not. Knowing which you have decides how much a comparison can be trusted.
Some Arrive In Hours Rather Than Weeks
Certain demand appears within hours of a triggering event. A storm, a cold snap, a burst pipe across a town. A monthly report cannot see any of it, which means the reporting cadence itself is wrong for those businesses.
What that demand looks like. A vertical line.
Nothing, then a great deal, then nothing again. Averaged into a month it disappears entirely into the surrounding weeks.
Why a monthly view fails it. The event is shorter than the period.
A month containing one intense week and three quiet ones reports as an ordinary month. The most important thing that happened is invisible.
What it means commercially. Readiness matters more than reach.
For these businesses, being findable at the moment demand appears is worth more than steady visibility. That changes what the work should aim at.
The reporting consequence. Look weekly.
Not because more frequent reporting is generally better. Because a monthly window cannot contain the event. This is the one case where a shorter view is genuinely required.
What to record. The triggers.
A dated note of the events themselves. Six months later, a spike with no explanation attached looks like an anomaly rather than a pattern.
The reverse risk. An unusual year.
A mild winter or a dry spring simply does not produce the demand. That is not the work failing and it will look identical in a report.
Two Seasons Pulling Opposite Ways
Some businesses have opposing peaks in the same year. One service busy in winter, another busy in summer. The annual total flattens out and both patterns become invisible, which is how a serious decline in one half goes unnoticed for a year.
How it happens. A broad service range.
Most established trades add services over time. Those services rarely share a rhythm. The wider the range, the more likely the peaks oppose.
What the total shows. Nothing at all.
Steady demand month after month, which looks like a stable business. Underneath, two things are moving in opposite directions.
Why that is dangerous. Cancellation hides decline.
If one service is genuinely losing ground while the other grows, the total conceals it completely. Nobody investigates a flat line.
The commercial version. Unequal value.
The two halves rarely earn the same. A flat total can hide profitable work falling away while lower value work replaces it.
How to see it. Split by service.
Segmenting by what the pages are about is the only way this becomes visible, which is why an unsegmented total is close to useless for a business like this.
What it changes about planning. Two calendars.
Each half needs building ahead of its own season, which means the publishing plan has two peaks rather than one.
Search Demand Peaks Before The Season
People search in advance of buying. That means the search peak arrives before the trading peak. Comparing search data against sales data without allowing for the gap produces a conclusion that is confidently wrong in both directions.
Why the gap exists. Deciding takes time.
Somebody researches, compares, gets quotes and then commits. All the searching happens at the start and all the money arrives at the end.
What that looks like in a report. Mismatched curves.
Search interest rising while sales are still flat, then search interest falling while the phone is at its busiest. Both look alarming and both are normal.
The first wrong conclusion. Early panic.
Sales have not moved yet, so the work looks ineffective. In fact the demand is building and has not converted.
The second wrong conclusion. Late panic.
Search traffic falls while the business is at its busiest, which reads as a collapse. It is the season moving from researching to buying.
How long the gap is. We will not say.
It differs enormously by trade, by value and by how urgent the need is. Any figure we gave would be wrong for most readers.
What to do instead. Align the comparison.
Compare search data against search data and sales against sales. Setting one against the other across a period requires knowing your own gap, which only your records can tell you.
The useful consequence. An early warning.
Search demand running ahead of trading means it can be watched as a forecast. A weak search season predicts a weak trading season, in time to do something about it.
Moving Dates Break Year On Year
Some annual peaks fall on a different date each year. When that happens, the same calendar month contains the peak one year and misses it the next, which breaks a year on year comparison in a way that looks exactly like a real change.
Which dates move. More than you would think.
Easter, school terms, bank holidays that shift by a weekend, anything tied to the tax year end and any event scheduled by an organisation rather than by the calendar.
What the comparison shows. A collapse or a surge.
A month that held the peak last year and does not this year appears to have lost enormously. The peak simply moved into the next month.
Why it is so convincing. The figure is real.
Nothing is wrong with the data. The two periods being compared are genuinely different in a way nobody labelled.
The fix. Align to the event.
Compare the weeks around the event rather than the calendar month containing it. That restores the like for like comparison immediately.
The other fix. Widen the window.
Comparing quarters rather than months absorbs a shift of a week or two, which is one more reason the quarter is the useful unit.
Where the comparison argument sits. Its own guide.
Month on month against year on year covers which comparison answers which question.
Establish The Shape Before You Judge Anything
Until you know your own pattern, no period can be read correctly. That makes establishing the shape part of setting a baseline rather than something to work out later when a figure moves unexpectedly.
What you need. A year of data. Or your own knowledge.
A year of your own data shows it directly. Failing that, somebody who has run the business for years knows the quiet period even without the data.
Why the second is legitimate. It is evidence.
Trade knowledge is not a substitute for measurement and it is a great deal better than assuming no pattern exists. Write it down before the reporting starts.
What to record. Four things.
When demand rises, when it falls, what drives it and whether the driving date moves. Four lines, which make every subsequent report readable.
Where this belongs. With the baseline.
Alongside everything else recorded before work begins, which our benchmarking guide covers.
What it prevents. The annual argument.
A shape agreed in advance ends the conversation about whether a quiet month is a problem, because both sides already know the answer.
When to revisit it. After anything structural.
A new service, a new area or a change of market can alter the shape. It is not a permanent record.
Do Not Let It Become An Excuse
Everything above makes seasonality a powerful explanation, which is exactly what makes it a convenient one. A supplier reaching for it every time a number falls is avoiding a question. A client should be able to tell the difference.
The test. Does it work both ways.
If seasonality explains every fall and never explains a rise, it is being used selectively. A genuine seasonal account credits the season for good months too.
The second test. Was it predicted.
A shape established in advance means a quiet period was expected before it arrived. Seasonality produced after the fact to explain a surprise is a different thing entirely.
The third test. Does the comparison support it.
If the same period last year was materially stronger, the season is not the explanation. That check takes minutes and settles it.
What seasonality cannot explain. Several things.
A fall concentrated in one section, a sudden drop on a specific date or a decline that does not match last year's pattern. Those need proper diagnosis.
What to ask. Two questions.
Did we expect this and does last year show the same shape. Both are answerable and neither is confrontational.
The reasonable position. Both can be true.
A season can be genuinely quiet and something can also be wrong. Accepting the first does not require ignoring the second.
Where everything sits. The hub.
All nineteen guides are on the measuring performance guide.
Search demand
peaks before
the trading
peak.
People research, compare and get quotes before they commit, so all the searching happens at the start and all the money arrives at the end. Search interest rises while sales are still flat, then falls while the phone is at its busiest. Both look alarming in a report and both are entirely normal, which is how the same business gets told twice a year that something has gone wrong.
What we establish before reading anything:
If seasonality explains every fall and never explains a rise, it is being used selectively. Ask whether the quiet period was predicted before it arrived.
Every guide.
One question.
What each number actually means, how to read the data without being misled, what belongs in a report and what does not, then how to connect any of it to the enquiries the business is actually paying for.
