What December Told You About the Year

Peaks reveal rather than distort

The usual instinct is to treat December as an anomaly to be excluded from any serious reading.

The totals are indeed unrepresentative and the ratios are not.

A conversion rate, a bounce pattern, or a checkout drop-off measured across a thousand visitors in December is a far better sample than the same measurement across two hundred in a normal month.

So December is the month with enough data to see things that are invisible the rest of the year.

The five ratios worth reading

• Enquiry or order rate per visitor.
• Mobile against desktop conversion.
• Where people left the checkout.
• Which source converted best.
• Which pages preceded a sale.

The second is the one that produces the most useful surprise. A mobile conversion rate well below desktop is present all year and is only visible with enough volume to be sure it is real rather than noise.

Compare December to December

Which is the only fair comparison.

Against November, December always looks better and tells you nothing.

Against last December, it tells you whether the business grew, whether the site improved, and whether last year's changes did anything.

Look at the same window in both years, allowing for the days of the week falling differently.

And write down the comparison, since next year you will want three points rather than two.

Two years of December data is the beginning of a trend, and one year is simply a number.

Record it either way, since the second year is worthless without the first.

A worked example

A shop read its December figures for ratios rather than totals.

Desktop converted at roughly three times the mobile rate, on similar traffic volumes.

They had assumed mobile was weaker and had never had enough data to know by how much.

Checking the checkout on a phone found a shipping selector that required a precise tap and a form field that triggered the wrong keyboard.

Both of those were fixed in January and the gap had narrowed measurably by the following December.

Neither problem was new; the December volume was simply what made it visible.

They had both been costing something quietly all year.

Look at where the traffic came from

Which changes in December and tells you something.

A month with unusual traffic composition shows which sources scale and which do not.

Search traffic that rises with the season is a source that responds to demand.

A source that stayed flat while everything else doubled is one that has a ceiling, which is worth knowing before investing more in it.

And a source producing volume with no conversions is one to reconsider entirely.

That comparison is only possible in a month where the totals move enough to see the difference.

Read the search terms too

Which is the free research December provides.

The queries bringing people to your site in a buying month are the ones with intent behind them.

Look at what people actually typed, particularly anything you had not thought of, and particularly anything phrased as a question.

Each of those is a page you could write in January, based on evidence rather than guessing.

The same applies to your own site's search box, if you have one, where the terms with no results are the clearest instruction you will get.

Note what broke under load

Which is the other thing a peak provides.

A slow page, a form that failed, an email that did not send, a stock count that went wrong.

Write each one down as it happens rather than trusting memory, since by January the details will have gone.

Anything that broke at three times normal volume will break again next year at the same point.

That list is easily the most valuable output of the month and it takes no analysis at all.

Do the reading in early January

Which is when it is possible and useful.

Not in December, when there is no time and the month is incomplete.

Not in March, when the detail has gone and the urgency with it.

The first quiet week of January, with an hour set aside, is the right moment, and it sets the priorities for the rest of the year.

Put that hour in the calendar now, since January fills up on its own without any help from anybody.

An unbooked intention loses to a booked meeting every time.

Watch how long people took to decide

Which is a December-specific reading with a use beyond the season.

Gift buyers decide faster than people buying for themselves, and the gap between first visit and purchase compresses noticeably in the final fortnight.

Looking at how many visits preceded a sale, and over how many days, tells you what your site has to accomplish in a single session when somebody is in a hurry.

A site that needs three visits to convert in March is losing the December buyer entirely, and the December figures are what make that visible.

The fix is usually that the price, the delivery timing, and the way to proceed all need to be answerable without leaving the page.

The counter-case

December is genuinely different.

Visitors are buying for other people, in a hurry, with different price sensitivity, so conclusions about behaviour do not transfer wholesale to March.

Some businesses also have no December peak, in which case their revealing month is a different one and the same reading applies to it.

And a shop with three hundred December visitors has better volume than usual and still not enough for confident conclusions.

Treat what you find as a hypothesis to check next year rather than as a settled finding.

Two Decembers pointing the same way is what makes it real.

Read the ratios rather than the totals, compare December to December, note what broke, and do it in the first week of January.

What to do

1. Read ratios, not totals.
2. Compare mobile to desktop.
3. Find where people left.
4. Compare to last December.
5. Read the search terms.
6. List what broke.
7. Book the hour in January.

Step two produces the most common useful finding, since a mobile conversion rate well below desktop is present all year and only becomes provable at December volumes.

The full-year version is covered in a year of analytics read once.

Frequently asked questions

West Coast Media Solutions Inc. provides web design, web development, hosting, digital marketing, and business consulting to organisations across Canada, drawing on more than twenty-five years in the field.