Month on Month vs Year on Year SEO Comparison: Which Matters More?
The single most practical reporting decision anybody makes, almost always made by default rather than deliberately. Reports are monthly because fees are monthly, so the comparison follows the invoice rather than the question.
The Default Is The Wrong One
Monthly reporting produces monthly comparison out of habit. Nobody chose it, nobody argued for it and it appears in almost every report in this industry. That is a formatting decision being mistaken for an analytical one. It changes what businesses conclude about their own performance.
Where the habit came from. Invoicing.
Fees are monthly, so reports are monthly, so the comparison became monthly. None of that reasoning involved anybody asking what the comparison was supposed to show.
Why it goes unquestioned. It looks obvious.
Comparing this month against last month feels like the natural thing to do. Anything that feels obvious rarely gets examined, which is precisely why this one persists.
Who it affects most. Seasonal businesses.
Nearly every trade we work with has a shape to its year. In one of those, adjacent months are the two least comparable periods available.
What the alternative changes. The conclusion.
The same underlying performance can look like growth or decline depending purely on which comparison a report uses. Nothing about the business differs between the two readings.
What this page argues. Both, deliberately.
Not that one is correct and the other wrong. That each answers a different question. A report should choose rather than default.
What Month On Month Actually Tells You
Monthly comparison has two genuine jobs and it does both of them well. The trouble starts when it is asked to do a third thing, which is describing whether the business is better off, because that is not something it can answer.
The first job. Spotting a break.
Something failing shows up immediately when this month is set against last. A tracking fault, a page that stopped working or a technical problem all appear in that comparison first.
Why it is good at that. Speed.
You want to know about a break quickly rather than accurately. Monthly comparison is the fastest signal available. For that purpose speed matters more than precision.
The second job. Feedback on a specific change.
Something was altered last month and you want to know whether anything moved. That is a narrow question about a short period, which suits a short comparison.
The condition attached. One change at a time.
The feedback only means anything if you know what was changed. Several alterations in a month leaves you with movement and no attribution.
What it cannot do. Answer the real question.
Whether the business is better off than it was. That needs a comparison unaffected by the calendar, which is block four.
How to use it properly. As a monitor.
Treat it as an alarm rather than a verdict. Something moved sharply, go and look. Nothing more than that.
Why It Misleads
Every problem with monthly comparison comes from the calendar rather than from the business. Four of them, any one of which moves a figure noticeably while nothing about the company has changed.
Different lengths. The obvious one.
Months vary by several days. A shorter month loses that traffic before anything else happens, which produces a decline nobody caused.
Different trading days. Less obvious and larger.
Two months of equal length can hold different numbers of weekends. For a business whose enquiries arrive on weekdays, that shifts the figure more than the length does.
Holidays. Regularly ignored.
Bank holidays, school breaks and the period around Christmas change behaviour for everybody. A comparison spanning one is comparing two different kinds of month.
Seasonal position. The largest of all.
In a seasonal trade, a strong month can follow a stronger one and read as failure. The business had an excellent two months and the report says it declined.
Why that last one matters most here. Our clients.
Most of the businesses in this programme have a documented shape to their year, which our seasonality guide covers.
What all four share. Nothing to do with you.
None reflects anything about the work, the site or the market. All four are artefacts of dividing a year into unequal named blocks.
What Year On Year Gives You
Comparing the same month against itself a year earlier removes the length problem, the holiday problem and the seasonal problem in a single step. That is why it is the comparison that answers whether the business is better off than it was.
Why it works. Everything is held steady.
Same month, so the same length. Same position in the year, so the same season. Usually the same holidays. Only the business differs, which is what you wanted to measure.
What it therefore answers. The real question.
Are we better off. Not whether this month beat last month, which nobody outside a report actually cares about.
What it does to a seasonal trade. Makes it readable.
A quiet month compared against the same quiet month last year shows whether the quiet period is improving. That is invisible in any monthly comparison.
The trading days caveat. Not fully removed.
The same month in different years can still hold different weekend counts. Smaller than the other problems and worth knowing about.
Where it belongs in a report. The headline.
This is the comparison that should lead, since it answers the question the business is actually asking when it reads a report.
What it does not remove. Its own limitations.
Two of them, both real. Block five covers them.
Its Limitation
Year on year is the better comparison and it is not sufficient on its own. Two limitations, the second subtle enough that businesses regularly miss a developing problem because of it.
The first limitation. It needs a year.
A new site does not have one. Neither does a site rebuilt or migrated recently, since the data before the change describes a different website.
What to do without one. Say so.
Work with what exists and state the limitation plainly. A report that quietly compares against incomparable history is worse than one admitting it cannot.
The second limitation. It can hide a recent problem.
A year that started strongly carries the comparison. Something going wrong in the past few weeks stays invisible because the earlier months are still doing the work.
Why that is dangerous. It reassures.
The headline says you are ahead of last year while the current trajectory is downward. That is the worst combination available, since nobody investigates good news.
How to catch it. Keep the short view too.
The monthly comparison, used as an alarm rather than a verdict, catches exactly this. That is the practical case for keeping both.
The third thing worth noting. Last year may have been odd.
Comparing against an unusually poor period flatters everything. A comparison is only as good as the period it is measured against.
Use Both, For Different Questions
This was never a choice between two options. Each answers a different question, so a report should carry both and state plainly which one it is showing at any point.
The question year on year answers. Are we better off.
The commercial question, the one an owner is actually asking and the one that should lead any report.
The question month on month answers. Has anything broken.
An operational question, checked frequently and acted on quickly. It belongs in the report as a health check rather than as a result.
How to arrange them. Long first, short second.
Open with the year comparison because it answers the question. Follow with the monthly one because it catches what the year comparison hides.
What happens when they disagree. The useful moment.
Ahead of last year while falling against last month is a specific and important signal. Neither comparison alone produces it.
What not to do. Choose the flattering one.
Presenting whichever comparison looks better each month is the commonest sleight of hand in reporting. It is also obvious to anybody paying attention.
How this connects to the window question. Different subject.
Which comparison to use is separate from how long a period to read. Our trends guide covers the second.
Rolling Periods Are Better Than Calendar Months
A rolling period is a fixed span ending today, compared against the equivalent span immediately before it. It removes the month length problem entirely, it is easy to produce and almost nobody does it.
Why it removes the problem. Equal spans.
Both periods contain the same number of days by construction. Nothing about February being short or October being long enters the comparison at all.
What it does to trading days. Largely fixes them too.
A span covering whole weeks contains the same number of weekends on both sides. That removes the second calendar problem alongside the first.
Why it also reads better. No artificial reset.
Performance does not restart on the first of the month. A rolling view reflects that, while a calendar view imposes a boundary the business does not experience.
The practical advantage. It is always current.
You can read it on any day rather than waiting for a month to close. That matters when something is going wrong.
What it does not fix. Seasonality.
A rolling span against the one before it still crosses seasonal boundaries. For that you still want the year comparison, which is why rolling supplements rather than replaces.
Why almost nobody uses it. Reports are monthly.
The same habit as everything else on this page. Once a report is built around calendar months, a rolling view looks like extra work.
Label It Or It Will Be Misread
A percentage with no stated comparison basis is the commonest way a report misleads without anybody intending it. The figure is correct, the reader supplies an assumption and the two do not match.
What the reader assumes. Usually last month.
Absent any label, most people assume the most recent comparison. If the figure was against last year, they have just misread it in a specific and predictable way.
How the label should read. Beside the number.
Not in a footnote and not in a methodology section. Alongside the figure itself, where somebody skimming will see it.
Why it usually goes missing. Automation.
A tool produced the figure and nobody added the wording. That is not deception and the effect on the reader is identical.
What else needs labelling. The source.
Which system the figure came from, since search reporting and site analytics count different things and their totals will never match.
The question to ask. Compared to what.
If a report shows you a percentage and you cannot tell what it is measured against, asking is entirely reasonable. A good supplier welcomes it.
Where the reporting standard sits. Its own guide.
What belongs in a monthly report covers it. Everything sits on the measuring performance guide.
Your report
compares months
because your
invoice does.
Fees are monthly, so reports are monthly, so the comparison became monthly. Nobody chose it and nobody argued for it. In a seasonal trade that makes adjacent months the two least comparable periods available, so a strong month following a stronger one reads as failure while the business has just had an excellent quarter.
How we handle comparisons:
Ahead of last year while falling against last month is a specific and important signal. Neither comparison on its own produces it.
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.
