KPIs

Google Analytics events as KPIs

Measure a release with an event Google Analytics already tracks: when that beats Needlr's own counting, how to pick one, and its limits.

In Needlr: New check, step 2 (Before vs after)
0:00 / 1:10

Needlr can count a KPI itself, or read it from Google Analytics. Needlr's own count starts the day you publish. GA4 already holds months of history.

A Google Analytics event is something GA4 already counts on your site, like a lead, a sign-up or a purchase. You can use it as the KPI of a Before vs after impact check.

Think of a stopwatch and a sports watch

You got new running shoes in March. Did they make you faster? A stopwatch you buy today can't say. It has no times from before March.

Your sports watch logged every run all year, so it can. Needlr's own counting is the new stopwatch: it starts the day you publish a KPI. Google Analytics is the sports watch.

When to use it

Good for
  • A change that's already live for everyone, measured before vs after.
  • A KPI Needlr wasn't counting yet before the release.
  • A release from the past, up to a year ago (more).
Use something else when
  • Split tests. GA4 doesn't know which visitor saw which version (split tests).
  • Adoption. It counts people one by one, and GA4 only gives daily totals (adoption).
  • A KPI Needlr already counted for 4 weeks before the release. Its own counting works, and it also checks for unusual traffic. GA4 numbers don't get that check.

Before you start

Connect Google Analytics on your project's Settings tab (how). Without it, no GA4 events show up.

How to pick one

  1. 1Open your project and press + New impact check.
  2. 2In step 1, pick Before vs after and set the release date. The step is now called What did you ship?
  3. 3In step 2, Which needle should it move?, your GA4 key events appear next to your KPIs, each with a small GA4 tag. Up to 6 are shown.
  4. 4Click one. The line under it says Counted by Google Analytics · history available, results right away.
  5. 5Press Start measuring. Nothing is published to your site, so you don't need Tag Manager or an authenticator code.
Which needle should it move?
Signup finished
Demo booked
Generate leadGA4
PurchaseGA4
+ New KPI
Counted by Google Analytics · history available, results right away
Step 2 of a before vs after impact check, with Google Analytics connected.

Two more ways in

  • A page. Press + New KPI and choose Reach a page. With Google Analytics connected, Google Analytics counts it, with its history. Press Change to count it with Needlr instead.
  • A fallback. If the KPI you picked is new, Needlr has no history from before the release, and step 3 says so. It then offers Google Analytics events that do have history, with how often each happened in the last 28 days. Click one to use it.

Events GA4 records by itself on every site, like page_view or scroll, aren't offered as a fallback. They rarely mean anything for a release.

How Needlr measures it

Each day, Google Analytics reports how many people visited and how many did the event. Needlr compares the share of people who did it in the weeks after the release with the same number of weeks before it, up to 4 weeks on each side.

  • A release more than 4 weeks ago gets its verdict right away.
  • A more recent release shows Too early to call until 4 full weeks have passed.
  • Right after you start, you may see Collecting data for a moment while the first numbers arrive.

Limits to know

  • Before vs after only. Never split tests or adoption.
  • Up to a year back. The release date can be at most a year ago.
  • Freshness. Needlr refreshes the numbers every hour, because GA4 keeps adjusting the last two days. Refresh numbers now on the Google Analytics card fetches them at once.
  • People, not clicks. The KPI's details show how many people did it, not how many times, and no pages (KPI details).
  • No warning when it stops. Needlr can't see a GA4 event stop firing. If the numbers look odd, check the event in Google Analytics.
  • Can't be edited in Needlr. To measure something else, pick another event.
  • If you disconnect Google Analytics, impact checks that use it keep their last numbers but stop updating.

Common mistakes

  • Looking for GA4 events in a split test. They only appear once step 1 is set to Before vs after.
  • Picking an event that's too broad. An event that fires on every page tells you nothing about one change. Pick the result you want more of (what makes a good KPI).
  • Expecting a verdict the day after a release. The after needs 4 full weeks.
  • A property from another website. The GA4 property must be for the same website as the project.
GA4
Google Analytics 4, Google's tool that counts what people do on your website.
Key event
An event you marked as important in Google Analytics, like a lead or a purchase.
Daily totals
One number per day. GA4 tells Needlr how many people did something each day, not who they were.

What the video says

Needlr can count a KPI itself, or read it from Google Analytics. Needlr's own count starts the day you publish. GA4 already holds months of history.

Reach for a GA4 event when the release is already live for everyone, and Needlr wasn't counting that KPI before it. History turns a past release into a before vs after verdict.

Create a before vs after impact check. In step 2, your GA4 key events appear next to your KPIs with a small GA4 tag. Click one, and it is your needle.

A Reach a page KPI can be counted by GA4 too, with its history. And if your KPI has no past, step 3 suggests GA4 events that do.

GA4 only knows daily totals, not who saw which version. So GA4 events work for before vs after only, never for split tests or adoption, and the after needs four full weeks.

Needlr refreshes GA4 every hour, a release up to a year back is fair game, and nothing touches your site. For anything already shipped, your analytics history becomes the verdict.