The end of March once again saw the grumbles of a nation having to get up an hour earlier, and another hour or more figuring out how to correct car clocks, microwaves and various other devices. Daylight Savings Time pushes the UK forward an hour at the end of March, with a conciliatory return every October.
The effect of this – on two Sunday mornings each year, either 1am to 2am doesn’t happen at all (when we spring forward in March), or it happens twice (as we fall back in October).
For counting footfall, traffic, and so on we need to be careful that an hour’s worth of data isn’t lost as the clocks change and most systems seem to deal with this reasonably methodically – either smoothing out or splitting up the hour in some way. Still, places with a strong Night Time Economy could well see a blip on these nights that ought to be factored in.
The peculiarities of time & dates apply more broadly as well. Take Easter, which can fall anywhere within a four-and-a-half-week period. There is little doubt that an early Easter has a different impact in town centres to a late one, particularly if schools needs to break up their term times as a result.
Christmas is another key date in the calendar, of course, and the weekday on which it falls can make a difference as well.
2017 saw Christmas fall on a Monday. This had an impact on weekly reporting, since all the major shopping days fell in the previous week, so Week 51 looked particularly good this year just gone.
2016 saw Christmas fall on a Sunday, which shifted bank holidays about, so the Christmas bank holiday technically ended up after Boxing Day, on Tuesday 27th.
All of this will have implications for people taking leave – if Christmas falls midweek, for instance, workers might be tempted (or forced) to take a few extra days’ leave to save coming in on odd days over the period.
Even the very way we count weeks and months can bring their own headaches! Many of us use month on month reports for comparison, but look closely at the definition of a month – it’s not usually from the first to the last day of the month, but an approximate number of weeks.
For instance, I have just received a footfall report for March 2018 – but look closely and it’s actually showing figures from 26th Feb to 1 Apr.
This is to make a fair comparison – “here’s the science bit…”: if we took March 2018 as-is, it gave us 5 Saturdays, where March 2017 only had four, which would skew the figures. Hence, why the nearest five full weeks are used – it makes sure the number of days are consistent and fair.
This doesn’t fit neatly into a year though…, so retail calendars usually end up with a 4-4-5 pattern (or similar) – that is, four weeks in January; four weeks in February and five weeks in March – and the pattern repeats from there.
To make matters more complicated, years don’t divide into complete weeks either, with one or two extra days accumulating per year. Every so often, January becomes a 5-week month to account for this! Phew!
Which is the first day of the week? Most reports use Monday as the first day, but sales figures might run Sunday-Saturday in keeping with retail practices.
Even the week numbers themselves are a matter of interpretation. Retail calendars, which determine the 4-4-5 pattern and are used extensively in footfall, use a slightly different week numbering system to the ISO scheme, which is a separate international system used in diaries and electronic calendars.
So, Valentines Day 2015 was in Week 6 according to the retail calendar and – by extension – most footfall and sales reports, but both Outlook and your bog-standard stationery shop diary will say it was Week 7.
These various foibles, quirks and ‘gotchas’ in dates and time are well worth remembering, particularly when comparing long term. The retail calendar system is generally good for comparisons, is well-tested and is widely used by High Street businesses, but it won’t be the same system as a supermarket wall planner…
Easter, Christmas and other dates are worth keeping a close eye on, too, as the date they fall upon could make a big difference both in performance and year-on-year comparisons. This is important for commentary and analysis of events.
And as for Daylight Savings Time, whilst unlikely to make an impact for most, it is something to be aware of twice a year. For the most part though, I’m just looking forward to getting my hour of sleep back in October!
If you’re not getting the data you need, or you’re not getting what you need from your data, we’d be pleased to have a chat.