What is Baseline?
A baseline is the measurement you take before you change anything: the state before the relaunch, before the campaign, before the new form. It consists of a few deliberately chosen numbers and it carries a date. What matters is not how much you record but that it is held outside the tools, as a file or a note, so it never has to be reconstructed from memory later.
It is rarely missing through carelessness: the tools look as though they keep everything. They do not. Google Analytics retains event data for free-form analysis for 2 or 14 months depending on the setting, the field data history reaches back 40 weeks, Search Console 16 months. Anyone who wants to run the comparison in two years needs today's file, because by then the longest of those windows has closed as well.
A baseline includes the case count the value came from. With small counts the swing is wide: at 20 conversions a month a difference only separates from chance at roughly 86 per cent growth, while at 250 conversions 19 per cent is enough. A starting value without its case count is a number without a yardstick, and the later comparison turns into a matter of opinion.
Why does Baseline matter?
A baseline proves its worth precisely when it becomes inconvenient: it stops a result from being arranged after the fact. Without a captured starting value, memory decides the review, and memory almost always favours whoever has just built something. The work costs half an hour, and the window for it closes with the rebuild.
Baseline in practice
- 01Before the relaunch, twelve months of Search Console data per URL are exported and parked as a file, independent of any tool.
- 02Before the first ad campaign starts, the team notes how many enquiries came in during the three months before, with no ads running.
- 03Before a form is rebuilt, views, abandonments and submissions are counted for four weeks so the later comparison has a reference.
