Crate serp_ctr 

Search-console arithmetic for SERP performance data.

This crate does the small pile of arithmetic that sits between a Search Console export and a decision: click-through rate, impression-weighted average position, position movement between two periods, and a projection of clicks at a different position using a CTR-by-position curve.

What this crate deliberately does not ship

There is no built-in CTR curve. Published “average CTR by position” tables are third-party estimates that differ by study, by query intent and by SERP layout, and baking one in would turn somebody else’s sample into this crate’s constant. So CtrCurve is something you construct from your own measured data — for example the impressions and clicks your own Search Console reports at each position. A projection is then a statement about your own history, not about the internet.

Every function here is pure integer or f64 arithmetic. No I/O, no dependencies.

Example

use serp_ctr::{ctr, weighted_average_position, PositionRow};

let ctr = ctr(7, 1_000).unwrap();
assert!((ctr - 0.007).abs() < 1e-12);

let rows = vec![
    PositionRow { position: 3.0, impressions: 900 },
    PositionRow { position: 40.0, impressions: 100 },
];
let avg = weighted_average_position(&rows).unwrap();
assert!((avg - 6.7).abs() < 1e-12);

Built for the pipeline behind https://toolsthatrank.com/, which verifies a figure against its source before it ships one.

Structs

CtrCurve

A CTR-by-position curve built from observed data.

PositionRow

One row of a search-performance export: an average position and the impressions behind it.

Totals

Aggregate totals for a set of rows: clicks, impressions and the pooled CTR.

Enums

SerpError

Errors returned when an input cannot support the requested arithmetic.

Functions

ctr

Click-through rate as a fraction in 0.0..=1.0.

ctr_percent

Click-through rate expressed in percent, rounded to decimals places.

page_of_position

The SERP page a position falls on, given per_page results per page.

pooled_totals

Pool (clicks, impressions) rows into a single Totals.

position_gain

Change in position between two periods, positive when the position improved.

weighted_average_position

Impression-weighted average position across many rows.