Collection has a deadline around the middle of this year, reporting has until next summer, and exporting is whenever you choose. Doing them in that order spreads the work over a year.

One date, three deadlines

The announcement gives a single date, which makes the whole thing look like one task with fifteen months attached.

It is three tasks with three different timescales, and treating them as one is why people either panic or do nothing.

Collecting data in the new property has a deadline that is genuinely soon.

Switching your reporting over has until the middle of next year.

Exporting your existing history has no external deadline at all until near the end.

Separating those turns an intimidating project into three small ones, only one of which is urgent.

The one with a clock

Collection, and the reasoning is arithmetic rather than opinion.

Nothing transfers between the two systems, so the new property only knows what it has recorded itself.

Year-on-year comparison requires a year of data, which means comparing next spring against this one requires collection to have started by this spring.

And to have a full year of comparable figures covering the period before the old system stops on 1 July 2023, collection needs to be running by about July this year.

That is a few months away, and it is the only genuinely time-sensitive part of the whole exercise.

It also takes about an hour, which is the frustrating part: a short task with a real deadline, sitting behind a large task with a distant one.

The one with a long runway

Switching your actual reporting, which is where most of the perceived work lives.

Learning the interface, rebuilding the reports you use, and becoming confident enough to answer a follow-up question all take time and none of it is urgent.

You have until the middle of next year, and the old reports keep working throughout.

Which means an hour a month for the next year is more than enough, and considerably more effective than a day spent in a panic next June.

Anybody telling you this must be done now is describing the collection task and applying its urgency to the whole thing.

The one you set yourself

Exporting your history, which has no deadline until the final months.

Existing data stays accessible for a period after processing stops, so the practical window extends beyond next July.

The reason not to leave it to the end is that everybody else will, and because an afternoon in a quiet week is easier to find than an afternoon in a deadline month.

Pick a month with nothing else in it, do it once, and cross it off.

The one thing worth doing before then is a rough note of what you would want, so that when you sit down you are not deciding and exporting at the same time.

The order that works

Seven items across fifteen months, of which two are this month and neither takes more than an hour.

A worked example

A business responded to the announcement by booking a full day with a consultant for the following week.

The consultant asked what they wanted from it and the honest answer was to have the whole thing dealt with.

The work that genuinely could not wait took ninety minutes: a property created, tracking verified on interior pages, and two conversions configured.

The rest of the day would have been spent rebuilding reports the business would not look at for a year and would have forgotten how to reach.

They took the ninety minutes and a written plan with dates instead, and booked a second short session for the autumn.

By the following spring they were reporting from the new system, having spent perhaps five hours in total spread across a year.

Why spreading it works better

Not only because it is easier to fit in.

Learning an interface in one sitting produces knowledge that is gone within a month, because nothing has been used since.

Rebuilding one report a month means each is used before the next is attempted, which is how anything actually gets learned.

It also means that when a question arises, the new system has a partial answer available, which builds confidence gradually rather than demanding it all at once.

And it avoids the specific failure of migrating fully in a day, discovering three months later that something was configured wrongly, and having no historical familiarity to notice with.

Where the runway is shorter

Two situations deserve more urgency than the general case.

A business with strong seasonality, where the meaningful comparison is a specific month against the same month last year, has a harder constraint: to compare next January you need this January already recorded, which may already have passed.

And anybody who reports to somebody else, particularly a client or a funder, needs the parallel period to be long enough to explain the transition credibly, which means starting earlier than a business answering only to itself.

In both cases the response is the same: get collection running now, and let the rest follow the timetable above.

The counter-case

There is an argument for doing less than this, not more.

A business that has never made a decision on the strength of an analytics figure could reasonably set up collection, spend nothing further, and decide next year whether it wants any of it.

Some will conclude that a simpler tool, or counting enquiries properly, serves them better than either version of this product.

There is also a case for waiting on the reporting work specifically, since the interface is still changing and reports built now may be reorganised before you rely on them.

Collection is the part with a deadline. Everything else is a judgement about how much analytics your business actually uses.

The plan

  1. Separate the three tasks rather than treating them as one.
  2. Do collection now, in about an hour.
  3. Verify it on interior pages.
  4. Configure two conversions this week.
  5. Put the June check in the calendar.
  6. Rebuild one report a month from now.
  7. Export in a quiet week, not in the final month.

Step two is the only item on the list with a real deadline, and the deadline is months rather than the fifteen the announcement implies.

Why nothing carries across is covered in your historical data is not coming with you.

Frequently asked questions

How long do I actually have?

It depends which task. Collection has a deadline within months, switching your reporting has until the middle of next year, and exporting has no external deadline until near the end.

Why is collection the urgent part?

Because nothing transfers, so year-on-year comparison requires a year of recorded data. A full year covering the period before 1 July 2023 means collecting by about July this year.

How much time does the urgent part take?

About an hour: create the property, verify tracking on interior pages, and configure two or three conversions. The frustrating part is that it is short and has a real deadline.

Should I rebuild all my reports now?

No. An hour a month for a year is more effective than a day in a panic, because each report gets used before the next is attempted, which is how anything is learned.

When should I export my history?

In any quiet week over the next year. Existing data stays accessible for a period after processing stops, so the window extends beyond the date, but everybody will leave it to the end.

Who should move faster?

Businesses with strong seasonality, where comparing a specific month requires that month already recorded, and anybody reporting to a client or funder who needs a credible parallel period.

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.

Feeling behind on this?

Separate collection from reporting. The first takes an hour and is urgent; the second has a year and is what looked intimidating.

Start a Conversation