# Key Concepts

Before running experiments, understand these core concepts.

***

#### Hypothesis

A hypothesis is a specific, testable prediction about what change will improve a metric.

**Good hypothesis:** "Changing the CTA from 'Learn More' to 'Get Started Free' will increase button clicks by 15% because it reduces perceived commitment."

**Bad hypothesis:** "Let's try a different button."

A strong hypothesis includes:

* What you're changing
* What metric will improve
* Why you believe this will work

***

#### Variants

Variants are different versions of your page that you test against each other.

* **Control:** Your original page (unchanged)
* **Variant A, B, C...:** Modified versions

In an A/B test, you have one control and one variant. In a multivariate test, you test multiple variants simultaneously.

***

#### Statistical Significance

Statistical significance tells you whether your results are real or just random chance.

Fibr calculates this automatically. When an experiment reaches statistical significance (typically 95% confidence), you can trust that the winning variant actually performs better.

**Before significance:** Results might flip. Don't make decisions yet.

**After significance:** Results are reliable. Roll out the winner.

***

#### Traffic Allocation

Traffic allocation determines what percentage of visitors see each variant.

**50/50 split:** Half see control, half see variant. Fastest results but higher risk.

**90/10 split:** Most visitors see control, small group sees variant. Safer but slower.

Start with 50/50 unless you're testing something risky.


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://support.fibr.ai/hypothesis-and-experimentation/overview/key-concepts.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.