What Needlr does
Needlr tells you if a change you made really moved the number you care about.
Your team ships every week: new copy, a shorter form, a feature your coding agent built. Then someone asks the hard question. Did any of it move the needle?
Needlr tells you if a change you made to your website really worked. It answers one question: did it move the needle?
The needle is the number you care about, like how many people book a demo or start a trial. We call it your KPI.
Think of a bake sale
This week you put up a bigger sign at your bake sale, and you sold more cookies. Was it the sign? Or was it just a sunny day?
Needlr is the friend who counts fairly. It makes sure the sign is the only difference. Then it tells you: the sign helped, the sign hurt, or it made no real difference.
Who it's for
- Product managers and founders who ship changes and need to know which ones paid off.
- Marketers who try new words, offers and pages.
- Teams with coding agents. An AI that writes code can ask Needlr what is worth building, and whether it worked (more).
Three ways to measure
Each question you ask is an impact check: one change, one KPI. There are three kinds:
- Split test: for a change you haven't shown to everyone yet. Some visitors see the new version, the rest see the old one, at the same time (more).
- Before vs after: for a change that is already live for everyone. Needlr compares the 4 weeks before it with the 4 weeks after it (more).
- Adoption: for a new feature. Needlr counts how many people use it, against a target you set, like 20% within 30 days (more).
Not sure which one fits? Which check should I use? decides it in one question.
What you get
When there's enough data, every impact check ends with a verdict. A verdict never changes after it's given.
The change moved your KPI the right way. Keep it.
The change hurt your KPI. Go back.
No real difference. Keep whichever is simpler.
Something is wrong with the counting. Needlr puts the impact check on hold and tells you what to fix (more).
Before that, it says Collecting data or Too early to call. Needlr doesn't guess early (why).
What Needlr needs from you
- Your website's address, in a project (how).
- Google Tag Manager, connected once. That's how Needlr shows changes and counts your KPI (how).
- Optional: Google Analytics. With it, you can also measure changes you released in the past.
What Needlr will never do
- Change your live site without you. Every change to your site needs a code from your authenticator app.
- Let a coding agent publish. Agents can ask and read verdicts, but they can't change your site.
- Collect personal data. No names, no emails, no IP addresses. Just a random visitor id, the page and the KPI.
What the video says
Your team ships every week: new copy, a shorter form, a feature your coding agent built. Then someone asks the hard question. Did any of it move the needle?
Needlr answers with an impact check. You pick the change and one KPI, a demo booked or a trial started. Needlr measures it and gives you a verdict you can defend.
For a change you haven't released, your impact check is a split test: half your visitors see it, half don't. Same weeks, same campaigns. So the difference is your change.
Already shipped it to everyone? Needlr compares the weeks before the release with the weeks after. Launched a feature? It counts how many people actually use it, against your target.
Early numbers swing. Needlr waits for enough visitors and a full week, then gives one verdict that won't flip: improved, got worse, or no clear effect. Until then, it says not yet.
That's Needlr: every release gets a business verdict, from your team or your agents. Stop arguing about what worked. Know it, and spend the next sprint on what moves the needle.