LoginSign Up
← Back to all sparks
R

rfm

ANALYTICS
Velocity0.0

R package rfm by rsquaredacademy — release notes from GitHub.

github.com/rsquaredacademy/rfm ↗

A customer segmentation package that went quiet for six years and returned with dependency hygiene

r-packagecustomer-analyticssegmentationdependenciesdormancy
◆Current state
rfm computes recency, frequency and monetary segmentation for customer analytics in R. The feature surface was set early: 0.1.0 shipped a Shiny app and customer-level input, 0.2.0 added default segments and median statistics, 0.2.1 added user-specified score thresholds and returnable plot objects. Then nothing for nearly six years. Version 0.4.0 in April 2026 fixes a missing-column error and a customer id fault, and moves plotly and gganimate from Suggests to Imports.
◆Where it's heading
The 0.4.0 release says more about maintenance posture than about product direction — the version jump past 0.3.x with only two bug fixes and a dependency reshuffle suggests a package being brought back to a releasable state rather than resuming development. Promoting plotly and gganimate to Imports makes the visualization stack mandatory, which is a heavier install in exchange for a simpler code path. The core RFM computation itself has not changed in this window.
◆Prediction
The entries show a package returning from dormancy rather than pursuing a roadmap, so further small fixes are more likely than new segmentation capability.

◆Recent moves

  1. 3mo ago

    rfm 0.4.0

    Ends a six-year gap with two bug fixes — a missing column error on customer-level data and a customer id fault — plus moving plotly and gganimate to Imports. A return-to-releasable-state release rather than a resumption of feature work.

    View source ↗
  2. 6y ago

    rfm 0.2.2

    An urgent patch preventing segments from being overwritten. The last release before the package went dormant for nearly six years.

    View source ↗
  3. 6y ago

    rfm 0.2.1

    Adds user-specified score thresholds and an option to return plot objects rather than only drawing them, alongside a frequency bin class and ggplot2 compatibility updates. Configurable thresholds matter here because segment boundaries are the main thing analysts want to control.

    View source ↗
  4. 7y ago

    rfm 0.2.0

    Adds default customer segments and median recency, frequency and monetary statistics, and exports the data behind the plots. Shipping default segments is what made the package usable without the caller defining a segmentation scheme first.

    View source ↗
  5. 8y ago

    rfm 0.1.1

    A patch release fixing the bundled Shiny app. Small follow-up to the interactive interface introduced days earlier.

    View source ↗
  6. 8y ago

    rfm 0.1.0

    The first tracked release, adding a Shiny app for interactive analysis and support for customer-level data as input. It establishes the dual identity — programmatic package and point-and-click tool — that the package has kept since.

    View source ↗