Project note, August 2026: This post documents the measurement layer added alongside SoCalNomad’s new social roundup format.
A new social format can look successful for the wrong reasons.
The first carousel may receive more likes because it featured a famous artist. A post may satisfy existing followers without reaching anyone new. A format can grow reach while sending no one to the site.
Without stored measurements, every explanation remains a story told after looking at the platform screen.
Build The Gauge Beside The Lever
The roundup carousel and social-insights collector were developed together.
The publisher records each post and its platform identifier. A daily job then captures available account and post metrics into a local history rather than leaving them trapped in transient social dashboards.
The initial data includes follower totals and accessible engagement signals such as likes, comments, and Facebook shares. Additional metrics can be incorporated when platform permissions allow them.
The important design decision is the history. A current follower count is a number. Repeated observations create a trend and let changes be associated with specific publication experiments.
Existing Followers And New Followers Are Different Audiences
A niche event slate may earn strong engagement from people who already follow the account. A broad, recognizable roundup may reach more non-followers but receive a lower engagement rate.
Neither result is automatically better. They answer different objectives:
- Does this serve the audience already here?
- Does this attract a new audience?
- Does either group continue to the calendar or newsletter?
Optimizing only for likes can reward familiarity. Optimizing only for reach can reward disposable exposure.
Define A Kill Decision
Measurement should make it possible to stop an experiment, not merely justify continuing it.
The project already had an example: an automated channel that produced nearly no traffic despite many posts. Recording outcomes earlier would have shortened that test.
For roundups, the useful review compares format, subject mix, posting window, engagement, reach when available, follower movement, and downstream traffic. A few posts are not enough to establish a law, but they are enough to begin replacing intuition with evidence.
Instrumentation does not tell the publication what to value. It tells us whether a chosen tactic is producing the behavior it was built to produce.
That gauge belongs in the first version, before visual polish makes the experiment emotionally expensive to abandon.