No data moves between the two systems. Set the new property up now so it begins accumulating, and export your existing history yourself so you own a copy independent of either.

Two products, not two versions

The common assumption is that the new analytics is an upgrade, and that switching will bring your figures across the way a new phone brings your photographs.

It does not. They are separate systems with different data models, collecting independently.

A new property starts empty on the day you create it, with no history at all.

Which means the useful comparison, this month against the same month last year, only becomes possible twelve months after you start collecting.

What that costs in practice

The second is the one small businesses feel most, since knowing whether a quiet February is normal requires a previous February to compare it with.

Which makes today the argument

The practical conclusion, and it is the whole reason this matters now rather than later.

Every month without the new property collecting is a month of comparison data you will never have.

A business that set it up when it launched already has a year and can make year-on-year comparisons today.

A business that starts this month can make them next March, and one that waits until the autumn cannot until the autumn after.

The cost of setting it up is an hour. The cost of waiting is measured in months of history that cannot be recovered afterwards.

That is true regardless of when or whether you eventually switch over, which is why it is worth doing before any decision about switching.

Export what you already have

The other half, and it is worth doing for its own reasons.

Your existing history lives in an account controlled by somebody else, subject to their retention settings and their product decisions.

Holding your own copy is ordinary prudence, in the same way as holding your own backups, and it is not specific to any migration.

Exports are available as spreadsheet files from the reporting interface, a report at a time, which is tedious and entirely adequate for a small business.

Do it once, properly, and you have a permanent record independent of any account, any supplier, and any future change.

What to export

Not everything, since most of it will never be looked at again.

Monthly totals for visitors and sessions, going back as far as you have, which gives you the trend line.

Traffic by source, monthly, which is the comparison that actually informs decisions.

Your top thirty or forty pages by traffic, with their figures, for a few representative periods.

Goal or conversion totals by month, which is the record of what the site produced.

And a note of anything unusual you already know about: a month with a campaign, an outage, a redesign.

Four exports and a paragraph, which is an afternoon and covers everything anybody will ask in five years.

A worked example

A business created the new property and, seeing it empty, assumed something had gone wrong and abandoned it.

Fourteen months later they returned to it, created it again, and started from zero for the second time.

Had the first attempt been left running, they would have had over a year of history and could have compared seasons immediately.

Instead they were back at the beginning, with the same wait ahead of them.

They also exported their existing monthly totals and source breakdowns to a spreadsheet, which took an afternoon and turned out to be the more valuable half.

That file has since answered several questions about earlier years that no interface could, because it sits in a folder they control rather than in an account.

Keep the export somewhere durable

A short note, since an export in a downloads folder is not a record.

Put it with your other business documents, named clearly, with the date it was produced and what period it covers.

A spreadsheet with a tab per report and a first tab explaining what each contains is enough, and takes ten minutes to organise.

Add a line to your annual notes saying where it is, since the person looking for it in three years may not be you.

And repeat the export once more before you eventually stop using the old system, whenever that turns out to be, so the record runs right to the end.

Some things are worth recording outside either

A broader point that the export makes obvious.

The figures that matter most to a small business are frequently ones no analytics system records: how many enquiries arrived, how many became customers, what each was worth, and where they said they heard about you.

Those live in an inbox, a diary, or a person's memory, and they survive no migration at all because they were never in a system to begin with.

A single sheet with a row per month, recording enquiries, jobs won, and the main sources, is worth more than any analytics history and takes a minute a month.

It also has the advantage of continuing uninterrupted through any change of tool, which is exactly what this article is about.

The counter-case

Historical analytics is worth less than the effort of preserving it suggests.

Most small businesses have never made a decision on the strength of a figure from three years ago, and a great deal of stored data is detail nobody will ever consult.

Year-on-year comparison is genuinely useful and is also achievable from a simple record of monthly enquiries, which many businesses keep anyway and which is more meaningful than visit counts.

There is also a limit to exporting: pulling forty reports across six years is a project, and four reports covering the trend is the sensible version.

Set the new property collecting today, export the handful of things that matter, and do not build an archive nobody will open.

The afternoon

  1. Create the new property today if you have not.
  2. Confirm it is collecting before assuming it works.
  3. Export monthly visitors as far back as you have.
  4. Export traffic by source, monthly.
  5. Export your top pages for a few periods.
  6. Export conversion totals by month.
  7. Store it with your business documents, dated.

Step one is the item with a clock attached, since the history you will want next year has to start being collected now.

What a year of figures can tell you is covered in a year of analytics, read once.


Frequently asked questions

Does my existing data transfer to the new property?

No. They are separate systems with different data models, collecting independently. A new property starts empty on the day you create it.

What does that actually cost me?

No year-on-year comparison until a year has passed, no seasonal baseline, no before-and-after for earlier changes, and nothing to check a surprising figure against.

Why set it up now rather than later?

Because every month without it collecting is a month of comparison data you will never have. Setting it up costs an hour; waiting costs months of history.

Should I export my existing history?

Yes, as ordinary prudence rather than for any migration. It lives in an account controlled by somebody else, subject to their retention settings and product decisions.

What is worth exporting?

Monthly visitors as far back as you have, traffic by source, your top thirty or forty pages for a few periods, and conversion totals by month. Four exports and a paragraph.

Where should the export live?

With your business documents, named and dated, with a tab explaining what each report contains. Not in a downloads folder, which is not a record.

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.

Not set the new property up yet?

Do it today. The year-on-year comparison you will want next March has to start collecting this month.

Start a Conversation