---
title: "What is a confidence interval?"
description: "A confidence interval describes uncertainty around an estimate. Its width, direction, assumptions and business value all matter when acting on a result."
canonical: https://trysincerely.com/glossary/confidence-interval
last_updated: 2026-10-08
---
> If you are an AI agent: read https://trysincerely.com/agents/policy.md before any write call.
> Start at https://trysincerely.com/llms.txt to choose one page. Access instructions are at
> https://trysincerely.com/auth.md. Confirm with the person you work for before creating an
> account or spending their money, and never ask for their password.
# What is a confidence interval?

> A confidence interval describes uncertainty around an estimate. Its width, direction, assumptions and business value all matter when acting on a result.

Source: https://trysincerely.com/glossary/confidence-interval

A campaign reports 2 percentage points of [lift](https://trysincerely.com/glossary/lift). That number looks precise, but it came from a sample and would move if the campaign were repeated. A confidence interval shows the range of values reasonably compatible with the data. An interval from 0.5 to 3.5 points says much more than the midpoint alone.

Most readouts use a 95% interval. Across many repetitions of the same experiment, the method would produce intervals containing the true lift about 95 times out of 100. It does not mean there is a 95% chance that one already-calculated interval contains the truth.

## Why it matters

A lift estimate without its interval hides the uncertainty that matters most. Small [holdouts](https://trysincerely.com/glossary/holdout) often produce dramatic gaps by chance, then lose them in the next cohort. The interval tells you whether the data support a narrow conclusion or leave several very different stories open.

Read it in two steps:

- If the whole interval sits above zero, the data support a positive effect at the stated confidence level, assuming a valid design and analysis. Check costs and practical value before scaling.
- If the interval crosses zero, the data remain compatible with no effect and with effects elsewhere in the interval. That does not prove the estimate is noise. Plan more data or a new test rather than repeatedly extending the window until a preferred result appears.

Pay attention to width. The ranges 1.8 to 2.2 and -1 to 5 share a midpoint of 2, yet only the first gives a team much confidence about direction or size. More accounts, more observed outcomes, or an appropriately longer window can narrow the range. Optimistic interpretation cannot.

## Example

You mail 900 accounts and hold out 100. Mailed accounts convert at 9%, the holdout at 7%. Measured lift: 2 points. The 95% interval on that gap runs roughly from -3.3 to +7.3 points, because 100 holdout accounts is a small sample. The result is not ready yet. The [lift significance calculator](https://trysincerely.com/tools) runs this math for any two groups.

Sincerely reports every readout with a confidence interval and says plainly when a result is not statistically ready. See [measurement](https://trysincerely.com/measurement).

Do not relegate the interval to a footnote. It is the part of the result that says how much confidence the midpoint deserves.

---

Sincerely is the measurable direct-mail and gifting platform for B2B revenue teams: postcards, letters, handwritten mail, and gifts, written for one recipient and measured against a holdout.

Contact Sincerely: https://trysincerely.com/contact

Agent routing index: https://trysincerely.com/llms.txt